I think there's a bit of a straw man going on with your first two points MaxWar - maybe it's just me, but I don't ever say "ok these are the values I want production set at, let's make it so."  Whenever I'm tweaking production, I'm trying to optimize output: if I need exactly 28 science output from a particular planet to hit my next research this turn, I'm going to adjust the sliders and the wheel the same way - wherever I get 28 science, that's the number I want.  I'm entirely end-result focused, so being able to set to exactly 30% - or exactly 55.7143% - is entirely irrelevant to me.

I would be FAR more interested in a planetary wheel (and not the overall government wheel as that's a more strategic overview to civilization production) that displayed RESULTS.  I know, without having to think for even a second, that if I put the dot on the precise top edge of the wheel, I will be allocating 100% of the planet's production to cash - but I don't know what that cash value will be until I look at a different window.  If I could see at a glance, "ok this planet produces 55bc/turn at 100% economy, 44 hammers/turn at 100% manufacturing, and 12 beakers/turn at 100% science," - now THAT would be a system worth working in.  Moving a dot to see how many arbitrary numbers I am assigning to a thing is nowhere near as intuitive - or informative - as moving a dot that shows me how many production items I will get as a result.

In the case of an output-centered management screen, I'd have much better results with a wheel, since I can adjust any of the three values with equal effort - if the tooltip was dynamically updated, I wouldn't even have to let go of the mouse button until I was content with my assigned levels.

Neither do I.  The tests I made to come up with those results are indeed artificial tests. They do not necessarily reflect the way the controls are going to be used. 

The use of such tests is not to reflect reality but to set a controlled environment to isolate a specific aspect of the tools we are evaluating.

And I would bet that most of that improvement came from being able to get all three values to the approximate set of values you wanted as one motion, rather than representing any significant savings in the fine-tuning time. There's also a question of interface responsivity; GCII's sliders potentially may not have the ability to change values as rapidly as the sliders hidden behind the disk do; they certainly wouldn't if for some reason you were using the arrow buttons rather than moving the slider bar.

Also, as said above, it's a largely pointless test. Are you ever really going to care that the production setting is 57/22/21 rather than 60/20/20 unless you're either being extremely anal or it marks the break-point on getting 22 hammers versus 23 hammers?

I also agree with iamderp that being able to set the output goal could be a more useful set of inputs than the split percentage. If I'm trying to balance the planetary budget, I don't really care if that comes at 57% wealth or 10% wealth, I just care that the net wealth production is 0. If I just want the purse world to be able to upgrade 1 economic improvement per X turns and send the rest into wealth, then I only really care about the number of hammers built, not whether it's 3% of capacity or 99% of capacity. I'm a bit more vague on when being able to set the exact research production would be useful, as I can't recall the research point cost of the techs being displayed in game and I'm not going to have a big list so that I can optimize things to get Tech A in exactly B turns of research, nor am I going to go through several of my planets setting sliders to optimize things in that way.

Yes, the GCII sliders can be locked. This does not mean that any arbitrary set of sliders can be locked, and for some strange reason not all functional and useful interface components are carried over from one piece of software to the next even when the general interface remains the same.

Yes,  you could say that. To be more precise, all symmetrical values are limited to 3 axis so once you know on which axis the value you need is, you only have one dimension to consider while moving the dot and it is easy to get there quickly. Most of the time consumed comes from dialing exact values. If aproximate is good enough for you it is even faster.

It is hard to make this affirmation without actual statistically valid data on how players make use of the controls. Some players have manifested that this degree of precision is indeed important to them.

Of course, real time monitoring of all parameters and various queue ETA would be requirement to use the wheel to its full potential. This is pretty much what I wanted to say in post #6 of this thread.

Please refer to the Orginal post and the question asked in the poll. The subject of this thread is to compare specifically the systems in GalCiv III and GalCiv II, not to some hypothetical system where the sliders cannot be locked.