Regarding the tech trees: I see a lot of people who assume that there is some kind of 'specialization' mechanic, but there's no statement to that effect -- just a simple comment that researching everything will leave you as a jack of all trades, master of none. And that that status will hurt.
This effect could also easily be achieved by simply removing the concept of 'diminishing returns' from the tech tree -- usually embodied as the concept that each level of tech costs more than the level that came before it. Or, better yet, have a 'growing returns' mechanic, where the delta of the effectiveness between each level gets stronger as you go further up the tree. Suddenly, the choice between grabbing diplomacy 1 or warmongerer 2 becomes a lot more interesting, and the further you go up the warmongerer tree, the less diplo 1 (or 2) looks like a good idea.
Lets wait and see how they're going to impliment things before we make decisions based on partial information.
Quoting Rhonin_the_wizard,
“The sales of GalCiv 2 were actually higher last year than they had been in the year before. We've put some considerable effort in trying to understand that. That’s not how it's supposed to work, if you understand.”
If a game is good it will continue to sell. It could also be that the game served a particular part of the market, and no other game did, or at least not as good.
I think the confusion point is that the rate at which the game was selling went up, which is extremely counter-intuitivie after years and years of sales. The market should have been saturated, with everyone who would have been interested having already bought it. (The rate at which 'new' 4X players 'evolve' into the market should also be static, unless something weird is going on). Personally, I wonder if the announcement of GalCiv3 had the same effect as an advertising campaign.