I am cautiously optimistic about GalCiv3, having enjoyed the previous two titles. However, I am concerned about the general trend of feature creep these days in game design. Some strategy games try so hard to please all of their fans and add feature after feature, with sprawling tech trees, fully customizable ships with dozens of components, RPG elements of heroes and leveling up ships/commanders. These is a lot of emphasis on freedom and allowing the player to do whatever he wants. This is often justified with the logic that more choices are inherently better. But really, are more choices necessarily better, or "deeper"?
I am wary because inevitably a lot of these so-called choices are merely illusions; dozens of the options are frankly not worth pursing, and only a few builds are acceptable. That a game designer can't make dozens of choices viable is not really his fault, since this is a superhuman task that is just not realistic to complete. Rather, the sin is in the attempt to please everyone by keeping meaningless choices as part of the game.
Take a game like Diablo 3, which pre-release was advertised to contain millions of possible build combinations. It turns out there's really only a few builds for each class that are viable. So millions of those so-called builds were so inefficient that players never ended up using them.
Players often demand to play the game their way. They also want to have success regardless of their choices. But if both of this is true, if all choices lead to the same outcome, then it just means that the choices are merely cosmetic. Choices that are merely cosmetic don't result in a strategy game. For a RPG, I don't think there's anything wrong with a game with only a few classes, as long as the playstyles of each class are sufficiently distinct. Keep the system simple, and make the choices meaningful.
I very well may be offending the "grognards" of the 4x crowd, but I don't think there's anything inherently interesting about being ordering your ships into such and such a formation, or firing their weapons at such and such range. I certainly don't imagine myself as some sort of Ender Wiggins, nor do I have any interest in that role. Designing a combat system with many options alone doesn't make the combat deep. Games like chess, bridge, or poker are fairly simple, yet they are very hard to master.
I'd rather see no tactical combat than a tactical combat system done poorly. My personal preference is no tactical combat, since in games of this type the mid-late game inevitably becomes bogged down in these mandatory battles. Auto-resolve is often cited as a choice for players who don't want to do the tactical battles. But is that really a choice? If the outcome is always worse, which is the standard for this type of game, then it's merely an illusion of choice. Who wants to intentionally gimp themselves by having a worse result? And needless to say, if the outcome is always better, no one would play the tactical battles anyway. The real reason to include auto-resolve is when the tactical battles cease to matter. This is usually because one fleet massively outclasses another. But by that point, the player has already won the strategic game, so, it's just checkmate in x moves. Not really interesting.
There's always the temptation to add more and more. More ships, more weapons, more tactics, more planets, more aliens. Perhaps it might be wise to, on occasion, step back and just keep the game's choices simple, but meaningful.
I'd much rather see GalCiv3 overcome the problem that has plagued all 4x strategy games: the problem of the snowball effect of exponential growth, such that the game is won in the early/mid period and late game is a foregone conclusion. Each game period should be meaningful and challenging. But this problem is not easily overcome, and solving it would require much thought and hard work. I'm much more interested in whether the designers have put any thought into this area.