For the record, GC3 is a terrific game as of 1.01, and I have loads of fun playing it. There are still a few things that could be improved in my opinion, and I'll elobarate on a problem I see with game pacing currently. The manufacturing and research modifiers from the pacing options are applied additively to those from improvements, instead of multiplicatively as I intuitively thought they would be, and in my opinion should be.
Example: At the start of the game, the -50% for very slow research makes breakthoughs take three times as long as with the +50% of very fast research. So far, working as intended. However, in the early mid game you'll research on specialized planets: Let's say you have +250% research from improvements, easily doable on a normal sized planet with adjacencies. Now, at very fast research speed you get +300%, at very slow +200%, which means that breakthroughs on very slow take only 33% more time than on very fast.
The gameplay effect of this is that improvements are much more powerful, relatively speaking, on slower pacing options, and a skilled player using highly specialized planets can overcome the handicap of slow pacing much more efficiently than an AI using balanced planets.
If at all possible, I would ask Stardock to consider making both research and manufacturing pacing truly multiplicative, i.e. very slow research combined with +250% from improvements gives 0.5 * 3.5 = 1.75 = +75% research speed compared to no improvements at normal pacing.