I didn't play Elemental. One of the reasons being, the very bad reviews. And I even posted about it in my blog (twice), because it was Stardock's, fantasy and turn-based strategy. As I never played it I'll give you my feedback on other things you did GREAT (I'm sure you know, but it's always good to receive extra praise I think). I'll also give you a couple of personal advices on how to increase your game's quality.
What you did great and how to improve it further
The Galactic Civilizations series is/was the space 4X reference of the beginning of the millennium. No need to elaborate further. Congratulations, and THANK YOU for a series of great and memorable games.
The Sins of a Solar Empires series is one of best space 3D RTS experiences there is. Period. People crave and flock for news on Rebellion (I know it by experience in my blog).
I know this may sound a bit simplistic, and perhaps not very helpful but, if you want to understand what you need to do in Elemental to make it better just look at Sins and GalCiv2. Now a few humble pieces of advice from a veteran gamer and games critic (and remember I never played Elemental). There are obviously many more factors that need to be considered and made well to make a great game but I think these 2 are probably the most important ones.
Is your game fun to play? Does it offer enough problems for the player to solve?
1. I have come realize over the years, by experience and literature that the single most important factor in a game is the "fun factor" (pretty obvious alright). But what isn't so obvious is what makes a game fun. Well, the equation (for me) is simple: Pleasure + Surprises. Pleasure being sensory stuff, eye candy and sounds/music (no smell, tact and taste yet, but we'll get there 
) Surprises are the key though, and that's why you can have completely unappealing games to the eye and hear and still have a ball playing. And in my experience there are no good or bad surprises, all surprises are equally good in catching the player's attention and force him to react and be curious.
2. What's the "problem" your game needs players to solve? Whatever your game offers, to be qualified as such it needs to offer a problem (or collection of problems) for the player to solve. It doesn't need to be only puzzle-type of problems, maybe your game aims on dominance? (eliminate all opposition) Is it a quest system (need to figure out what you need to know and do?). Need to solve a mystery? Players need a reason to play, they need to feel they are progressing towards some resolve, because that's what we humans do, we like (or need) to solve problems.
Conclusion
Didn't play Elemental due to the very bad reviews (among other personal reasons). GalCiv and Sins are outstanding products (never question that for a second). Look at those again, and again to understand what you did well. Does your game offers enough surprises? Is the player curious on going some place or do something? Does your game offer enough problems for the player to solve? Riddles, quests, objectives?
Good luck on this new Elemental installment. I'll give it a shot when it comes out.
PS: Since it's obvious that you liked Master of Magic, Civilization, Master of Orion and developed Galactic Civilizations, Sins of a Solar Empire and Elemental I think one of the natural games for you to do next is a Master of Orion sequel. You know it, we know it, it's the natural thing to do.
Thanks for your time.
Adam
