This is an interesting idea from a game-design perspective that I don't think would feel right in the context in which you are putting it. It just doesn't make any sense that knowing one thing would preclude you from learning something else. This idea would be much better suited for the ideology trees imo.

Whenever a game gives me choices with replay value, I eventually challenge/amuse myself by exploring the lower bound of wimpiness, roughly:

• How weak can I be and still win?

This takes various forms, depending on the game genre.

• eCCG.  There are power cards, uber-strategies, dominant decks.  I go the other way: I like to assemble C/U decks of crap commons & uncommons, 2-3 ultra-crap rares, and no very rares, take on the cheese sheep's dominant decks in open competition, and win.  I enjoy needling them by saying (as I win): You have more Very Rares in my HQ city than I have in my entire collection.
• This actually takes some skill in combinatorics.  You must envision a clockwork engine or combo of harmonious parts all interacting seamlessly, and then live long enough to draw it and get it into play.  So your total endgame firepower can be highly non-linear.  Cheeseheads do not grasp this, and just howl as it surpasses them.
• RPG.  I suppose a similar dynamic could exist in magic-user or cleric spell choices.  In old AD&D (3rd Ed?), there were some "coaster" spells that no player in his right mind would ever choose to take ... until you re-read them with an open mind, ponder the possibilities, and realize that in the right situation, they'd have effect waaaay out of proportion to a mindless combat spell.  This suggests the wacky niche of local specialist tour guide, who rents his services to passing adventurer groups who need to cross some tricky region in the game world.  If they slog in with their studly (but cheesy) combat strength, they'll valiantly bog down.  If they follow you instead, you lead them through like it's nothing.
• Magic-user had a 1st level spell, Affect Normal Fires, that ... seems totally useless.  Every cheesy battlemage wants Magic Missile for a precious 1st level slot.  But carefully read its description and compute the volume of smoke you can get from it, imagine you're trying to sneak through an enemy line (with six mounted dudes pulling two wagons), maybe add Invisibility and some thief and ranger friends -- and it can be the linchpin of an entire guerilla campaign.  It's an enabling technology that gives you a totally new niche of operations.  In the right campaign, that can totally trump Yet Another Damage Spell.

For 4X games, I confess I don't bother handicapping myself, because it's soooo much fun to out-research, out-tech, out-build, and out-gun the union of all AIs, and just conquer them first.  But the idea of many techs, some wimpy can creep toward this ideal

• As of Beta 3, I feel that GC3's tech tree actually lacks interesting choices.  Everybody needs some approval, manuf, research, growth, etc.  So all of my races tend to follow the same mainline tech path, and play the same way.
• It's not clear that a true weenie rush gambit strategy, sacrificing approval + growth to get a fighting fleet early (and then an early declaration of war + whack-a-shipyard), will ever thrive enough to survive.  Since you can't actually invade until AoW, you can't reap the benefits of this gambit.  So it might be only mutual assured destruction, like two snapping turtles biting each others' necks so that neither of them can retreat: the third raccoon that stumbles over them will wait patiently, then eat them both for free.  (In GC3 terms, if A weenie-whacks B, then I outgrow them both and easily conquer both A and B together.)
• I defer judgment on the interactions between influence, diplomacy, trade, etc.
• My "wimpy" strategy isn't actually "wimpy all the time".  Rather, I am actually building a gigantic Beware-I-Live Rube Goldberg contraption that comprises 6-12 moving parts.  Each of those parts, by itself and on its own, can be wimpy, crap, coaster, useless, which no self-respecting cheesehead would ever include.  (Popping up one tier higher, the mythical Game Designer, mumbling and cackling at his dimly-backlit keyboard, would have envisioned exactly this tradeoff: that the crap pieces do fit together this way.  So in that sense, I'm actually playing the game the way it was meant to be played -- or at least down one of its explicitly intended paths.)  I accept that temporary crapness because I think I can survive the race to reach the jackpot of synergy at the end.  So there must be a jackpot (and a Designer who put it there).
• Mapping this back to a 4X (such as GC3*, maybe with modded trees-of-buildables), it's not interesting to simply add some truly weak technologies to a tech tree.  There must be some pay-off at the end of a well-played path.  It might involve an intricate combination, such as:
• research this kind of outpost that annoys the space worms that infest your asteroid fields
• build that kind of ship with sensors that distinguish rock density (so you can tell worm poop from rock)
• buy up this civilian transport capability to haul things around within your ZOC
• harvest space worm poop as fertilizer that increases +food% to all planets in asteroid field effect
• spend your espionage points to steal Drengi ship models
• spend your xeno domestication points to habituate local space worms to come to your ships for free meals (which ships happen to look just like Drengi ships)
• finally, invest in mass drivers that sculpt asteroid orbits
• and then you can suddenly throw sixty asteroid-sized egg cases into Drengi space that hatch into fleets of space worms that seek out Drengi ships and eat metal.  Something totally off-the-wall and completely unbeatable like that.  (This is only one example )

I like the idea of totally weird tech paths that enable new things you can't do in any other way.  We could even give a broad palette of dozens of these techs, so that in any game you could only hope to explore a few of them.  (But that assumes a game engine that really does implement hundreds of capabilities under the hood, which may exceed Stardock's resources.)  Then evil Designers could hide many totally different jackpots among the combinations-of-techs, and let players engage in a cosmic meta-level Easter Egg hunt.  That would be a hoot.

A practical result of this line of thinking is that we can't just mod these combos ourselves, because mods are capped by the richness of the underlying game engine.  So maybe we can gently nudge the devs toward expanding that palette.

If you're finding difficulty in prioritizing tech choices, it's probably a well designed game.

Obvious choices should not be choices.

Agree that if you are having difficulty prioritizing that is a good thing. At the moment I am not having that difficulty. There's a fairly standard progression that makes sense and no real reason to change it. Not good.

+1 to the "Easter egg hunt". Occasional "holy crap that is cool and unique" technologies scattered through the tree would be great.

Regarding one tech choice not making sense to exclude another - think of it as sort of a research mindset. Something overlooked because energy am focus is elsewhere. Maybe you buy the other tech from someone else and all your researchers go "Ooooohhhh.... yeah, that makes sense but we were looking over here instead."

Finally on non-static tech trees - either SotS or SoaSE did that and it was fantastic. I mean truly awesome. You never knew what technologies you would have access to, particularly certain more far-out ones, or what your enemies might have. Huge game impact And would love to see that here.

Yeah, no, sorry.  For my taste the tech tree is too small and I've spent considerable time debating whether I am ambitious enough to put together a mod to make it 3-4 times larger and see just how flexible options on tech can be with it, like what Taslios said where techs are interconnected in interesting ways. 

Say, for example, you research holographics, star trek style, from force field technology.  At first you can entertain with it, holodecks and whatnot to raise approval, but then you can also make holographic factories that can retool in an instant to improve production or holographic miners to gather resources faster at mining starbases or holographic soldiers that raise soldiering.  At the high end, Holographic ship parts can be used to dramatically reduce the mass of components.  These are just a few examples to get started and we could go much, much further. 

I would like a larger tech tree as well.  And I don't really like the tech ages, they seem completely arbitrary to me.

I'll be brutally honest. I just want as many tech entries as humanly possible. When I reach the end of the tree, the incentive to play is somehow lost. So the more, the merrier. Never mind what they do, just crank 'em out!

What is SotS?

There is something to be said for the strat of gaming the game...  but how about a cool middle ground...

Velcro, Penicillin, Post it notes, Vulcanized Rubber, x-ray film, Superglue, LEDs, Teflon, and many many other things were all discovered on accident.

I understand that it is way to late to make massive changes to the game, but I would really like SOME element of surprise, or connectedness in the game.  right now the research just feels dead.  it is there but it is not interesting.  it is like a necessary evil and not something that to me makes the game either more interesting or more fun.

The tech tree should be more intertwined and cross correlated.  Researching in Military should open up options in Civics,  Diplomacy should open up access to terror weapons etc etc etc.    Right now the four areas do not link, and force a separation that is NOT found in real science.

I would also love there to be 200 or so Random just for fun "breakthroughs"  that the player has a 0-20% chance or so of discovering at random.   This would then give a bit of a wow factor.   if they really wanted to add depth they could have a random breakthrough create a whole new chain..  you "discovered superglue"  so it has a chain "Universal Cement" (faster build times on planet)  and "universal Solvent" (faster mining)  and "Laminar bonded armor" (better armor) etc etc etc..

You cut out where I said I understood it was too late to make massive changes to the game.     I work for a software developer... I fully understand how complex "simple" changes can be.

That said you also made my suggestion more complex than I was meaning it to be. (albeit if they could do it your way that would be awesome)

What I was intending is that once you randomly discover Superglue... it creates a new branch in the tree that would then lead to random bonuses to other techs within that field.  The rest would be preset and linear not random and exhaustive.

They could also make a rule that limited the frequency of discovery of these random techs...   so if there are say 200 random techs and you can at maximum discover them every 3 or 4 tech discoveries... then at most only discover like 20-30 of them a game...     just with random probability it could be years of gaming before a player hit all of them.

As for the AI not being able to read an interconnected tech tree...  I doubt the AI reads the tech tree at all in the same way humans to.   The AI probably sees all available techs and all possible routes to the techs.  It then has a few major techs that it prioritizes   Get to X by turn Y if possible.  If not then do Z logic strings.    Making the tech trees intertwined does not make the game any more complex to the AI.

Adding random techs would, however in the programing they have to have a hierarchy weight or value on each technology already assigned.   it would be rather simple to create a cross check that would weigh the benefit of x new tech against the value of Y tech and pics the one that is the most valuable...

As to your comment that Stardock always separates the tech trees...    I've been playing Gal Civ since it first showed up on Windows.   This is the first time that the research has felt artificially separate.   the Research focus's should make the game more interesting, but they just feel like mini-speed bumps at the moment.    Yes, I know Beta.. yes I know we have not seen everything yet.   But there is not really a oooh I can't wait to get x cause in this situation it will be really cool... vs  yeah this game I will avoid that cause it will suck choice..  every tech just adds a bit as per the OP thoughts it does not feel fun.. just tedious..   and I am usually of the mindset that the more techs and the bigger the tree the better... so there is definitely something missing to me.

I just hope that this feeling of something missing is corrected once the official game is launched.   Maybe adding more of the unique wonders, or a few tweaks here or there will be all that it takes.  Random surprise techs would be cool, but I suspect it is as you say far far too late for them to put those in prior to an expansion.