Yeah, I think you are exactly right in your analysis. Although it would be really cool and it would finally get those pesky constructors of my back. The AI spam them like crazy and sets upp starbases all over the place, even when he has nothing to gain from it.

Before bringing in a verisimilitude/'realism' argument, please consider what it is that you're asking for. When you ask for the ability to relocate a starbase, you are not asking for something remotely comparable to shifting the orbit of the ISS slightly. You are talking about moving a massive structure a distance which is at minimum on the order of 1 AU and may well be several lightyears, and while the fact that shipyards can do it does help the case that starbases perhaps could do it, it is by no means guaranteed. We have no real idea of the relative sizes or structural integrity of shipyards and other starbases, nor do we know whether or not there is some limitation in GalCiv FTL technology that prevents sufficiently-large structures from being moved (granted, we did have Terror Stars in GCII, which were about as mobile as shipyards are in GCIII, but those were, according to the listed masses of the stations and terror stars, roughly a quarter the size of a fully-upgraded economic starbase, or about as large as a station with ~5 modules, and the hull form of a Terror Star looks much more suitable to surviving the stresses of moving the structure than any type of starbase in GCII or GCIII does).

When you ask for the ability for a starbase-mounted weapon system to hit targets at the edge of the starbase's area of effect, you are asking for an FTL-capable weapon system, one whose FTL drive is capable of speeds at least comparable to and possibly significantly greater than the speeds which can be attained by standard vessels. After all, in the second-worst case scenario, the station's sensor bubble and the station's area of effect have the same radius (7 tiles), so the weapon needs to be capable of an FTL speed roughly 7 times greater than the FTL speed that the target is capable of (the missile needs to be able to cross 7 tiles in roughly the time it takes the target to cross 1 in order for the station to hit the target at the edge of its area of effect; the worst-case scenario is worse, because a station without sensor upgrades or a spotter is not even able to see targets beyond 4 tiles away, but its area of effect extends out to at least 5 tiles from the station, indicating that the station can attack targets that it doesn't know are present). This drive, which may have to be significantly faster than the drives found on regular ships, also needs to be sufficiently cheap to be viable for use on disposable ordnance, sufficiently small for it to be reasonable for the missile to be essentially immune to antiship weapons, and sufficiently small for it to be reasonable for the station to launch enough that there is a reasonable chance of the weapons penetrating the missile defenses of the targets (i.e. the system needs to present a credible threat to the target). The support for this being possible is rather lacking within the game; Tiny-hulled vessels have utility issues due to how difficult it is to give them reasonable offensive and defensive capabilities while maintaining enough strategic mobility to be effective over any significant distance. When you ask that the station's interceptor complement be permitted to engage the target, this on the surface appears somewhat more reasonable ... but these are the same kinds of fighters that are carried in carrier modules, craft which are presumably not FTL-capable in the first place, or if they are FTL-capable, they're likely too slow to keep up with the fleet in the first place. These fighters would need to be capable of the same kinds of speeds as the missiles in order to intercept interlopers, which would appear to be inconsistent with the rest of the game - I certainly cannot make a tiny-hulled fighter with only the modules that a Drone Interceptor has that can attain an FTL speed upwards of seven times greater than the speed of my standard ships; given that it's a straight copy of a Drone Interceptor (or assault fighter or guardian fighter), it's unlikely to attain more than perhaps half or a third the speed of a mid-game mid-size heavy warship seeing as it lacks any hyperdrive modules, which is at best barely adequate for a station with unupgraded AOE and fully-upgraded sensors (the target crosses ~11 tiles in the time that the interceptor must cross ~5 tiles) under the assumption that the interceptors launch as soon as the target enters the station's sensor range. You might be able to loosen the speed requirement of the interceptors by assuming that the station always has some out on patrols in positions where it's practical for the interceptors to reach any threatened point on the boundary of the station's AOE in time to intercept an interloper, but we're still hitting the issue of fighters which shouldn't be FTL-capable, or at least not very FTL-capable, and which shouldn't have particularly great endurance having reasonable FTL capability, reasonable endurance, or both.

Giving military starbases something to make them more worthwhile is not a bad idea, but using a verisimilitude argument to justify what you've suggested doesn't hold up to even a cursory examination of what capabilities the suggestion entails when we look at the apparent capabilities of the setting. More justifiable would be something like the military starbase providing nearby colonies with N fighters (the station provides the necessary logistical/command/whatever support through abstracted-out whatever, not station-based fighters rush out to threatened colonies whenever an attack occurs), or military starbases provide nearby colonies with a bonus to planetary defense or resistance because of the logistical support the station offers, or a nonmilitary bonus where colonies under the influence of a military starbase see increased tourism income due to all the off-duty military personnel or increased morale due to the clear evidence of the government's interest in protecting the region, or an indirect military bonus of boosting the (military) manufacturing output of nearby colonies (or perhaps even creating a small amount of military manufacturing for nearby shipyards).

(Also, the "who in their right minds builds a base without being able to relocate it" bit? Just about everyone builds bases without being able to relocate them intact. History is far, far more filled with permanent, non-relocatable installations than with installations that can be relocated trivially, and this includes most modern military installations. You are not simply going to relocate that airfield. You might build a new one and relocate the planes and equipment, but that's not really the same thing; most of the structure of the installation is going to remain in place unless you decide that it's a good idea to take the time and effort to take the place apart.)

Additionally, "influence borders =  political boundaries" is something that I despise about GCII and GCIII, and despite the developers' claim that it is not the case that influence borders are political borders but you have in-game options to make that the case, the game does a poor job of supporting that claim; the game provides much more support for the claim that influence borders are political borders, or at the very least claimed political borders. I do not care how strong my influence is in a region, that my influence borders surround your homeworld does not make that region into my space. Yes, this is an extreme example, but it is also something which is possible under the influence mechanics, particularly on easier difficulties. Making it even more the case that influence borders are political boundaries is not something that I will support.

verisimilitude- wow, cool word

I don't care if a game mechanic has the appearance of being true/real, i just want it to be fun and have a purpose.

I made myself a little rework to military starbases as a mod that I enjoy.

1) A module chain that can increase planetary resistance, defense, and add defensive interceptors to planets in its influence.

2) Increase range modules, to help make #1 useful.

3) Increased health systems that cost durantium

4) Increased sensor systems that cost thulium

5) Increased STRATEGIC repair modules that cost thulium

Many of these cost not one, but two construction points, and often special resources to help balance it. Its been pretty nice. I still build Econ, Mining, and if I'm trying to win culture, I build Culture, but now I'm also building at least a couple of these. As the AI gets better, these systems do as well because I'll need them.

The defense modules for planets are quite modest I might add, trying to keep them balanced, and many of these require appropriate tech level.

I think military bases sending fighters, even tiny/small craft designs, and missile volleys to a battle nearby is not only perfectly plausible but a good way of projecting its power. It would also allow people to guard their transports and colony ships a more effectively while they are in friendly territory.

Perhaps with a max of 2 starbases contributing? 

To anyone throwing realism arguments out there, if you really think about what you are writing, planetary and starbase missiles are actually a more plausible way of firing large ordinance across vast distances of space at large targets, (certainly other planets or starbases) but we have to throw realism out because this is a game, and ships make games fun. 

The simple answer is that we don't know, aside from that the unmodified range of a missile weapon is 1100 unspecified range units.

We can make guesses based upon the indications we can see of the engagement ranges in the battle viewer; ships being as large as they are relative to one another when they begin firing upon one another would imply that the engagement ranges are on the order of tens to maybe low hundreds of kilometers. We can assume that the ships are not to scale but that the battle viewer displays the engagement in real time and go off of the travel time of beam weapons since we know the propagation rate of one of them (at least, assuming that the weapon called a 'laser' is actually a laser), which would indicate that standard beam range is 0.2 light-seconds (~60,000km), and this would imply that if the ranges are all on the same linear scale then maximum missile range is about 0.34 light-seconds (~100,000km). We can try making guesses at how long the battles really take and work out how far apart the ships are based on the resulting timescaling and the known propagation rate of beam weapons, but that runs into serious issues when you realize that every action taken by a ship has to take a sufficiently small amount of time that it can take all of its actions in the space of one week, and so a ship with 30 actions per turn has only 1/30 of a week to spend fighting an engagement that uses up one action whereas a ship with 1 action per week has a full week in which to fight that engagement (further complicated by the fact that some of this time most likely has to be spent closing to engagement range).

We know from here that the official minimum tile dimension is ~3 million kilometers and that the official maximum tile dimension is ~10 parsecs (~300 trillion kilometers). It might be possible for ships to engage one another at a distance of 1 tile at the very low end of officially-possible tile dimensions; 3 million kilometers is about 10 light-seconds, and if the projectile speeds listed for missiles and mass drivers are on the same scale as the propagation rate of beam weapons and that scale is linear, then mass drivers could cross that distance in ~50 seconds, missiles in ~100-200 seconds, and beams in ~10 seconds. However, the requisite accuracy in predicted target location begins to stretch credulity with mass drivers (you need to know where the target will be nearly a minute into the future in order to hit the target with a mass driver, assuming the speeds are on a linear scale and that scale is the same as used for beams), as does the required accuracy in aiming the beam emitter or mass driver (even the largest GCIII ships have a maximum dimension of only about a kilometer, and your aim only needs to be off by 1.7e-7 radians for a shot aimed at the center of a 1km target to miss the target entirely if you're shooting from 3 million kilometers away). As such, I find it extremely unlikely that the engagement ranges are anything like 1 tile even on the extreme low end of tile dimensions

For what it's worth, my guess is that maximum missile range is probably somewhere around some thousands to low tens of thousands of kilometers.

I'd like to see just ONE starbase and add rings as you want to them. All rings. Rather than have 3 different types just build one and keep building it with additional rings. 

Also as has been stated back in Beta, the HP's of a Starbase are woefully out of line compared to a huge hull Dreadnaught. I personally think starbases which are player or ai made should START at 500 and gain 500 hps per major ring added with Military rings giving more offensive power. 

I agree military starbases need more purpose than a place to store a defensive fleet, which they do very well. That is the reason those castles mentioned earlier (and towers as well) protected a region: available troops could work to repel raiders and had a secure location for rest, repair and resupply. With this in mind, I agree there needs to be a way to deploy against encroaching ships, but it should be ship based. It was the troops in the castle, not the castle itself that threatened invaders. Without troops a castle could be passed up just as our starbases are now.

A player selected yes/no box could allow the starbase to launch up to the max logistical-points fleet available from docked ships and the player is notified in the event list. You can then <go to> and make a decision based on the situation. This would happen with enemies and neutrals, but not friendlies or races with right of access. By checking you can keep from killing that incoming freighter a neutral is sending over for trade, or send a second fleet out as backup.

This would require a change in how the borders work, but it would mean you need to be proactive in your defense. Modules giving fleets in the zone of effect advantages and enemies disadvantages are good, but adding specialized weapons to reach out hexes away just seems a stretch.

I also think closed borders are a bad idea. What I would like to see, however, is the end of Starbases in other players' ZOC without either the players being at war or as part of an open borders/alliance/specific construction permission treaty. Already built bases would remain after treaty expiration, but you'd have to renew to build new ones or expand existing ones.

Does anyone still use the "All Starbases Mine" and "Military Mining" mods? They are a step in the right direction.

I like the Castle idea.  Halve the maintenance of ships in a military starbase's ZOC and add a defense buff of 25% or so- then there's enough incentive to build them and to garrison otherwise idle warships there.

I think starbase HP should be nerfed up as well, increase with modules, and increase at twice the rate for military starbases.  Maybe not so drasticly as mentioned in earlier posts though.