It's a really fun game, but it does feel like a beta that's like two or three months from release or more.
I wonder why they pushed it out in this condition?
I think there are several reasons for this, but first my take on the facts of the matter.
1) The game is basically stable on a machine that can run it. I have experienced the after save crash bug a couple of times, which is really not that bad because the save is successful and the game can then be reloaded. Other than that, I've only seen one crash bug that appeared to be a race condition error caused by my clicking around too much in between turns. Whatever. In terms of priority A1 bugs on machines that are up to snuff, the after save crash bug is the only one I've identified in 40 or so hours of play.
2) Yes, there are a lot of little errors, most of them balance issues or holes in the AI. This is hardly surprising for a game of CiV's complexity, even in a "clean" initial release version.
Why is the game in this condition (which I don't find particularly objectionable)? First, it is a question of hitting a release date. The date was set a long time ago, and the firm had a large financial interest in hitting it. Delays to game release these days can be poisonously expensive, especially for relatively small development studios like Firaxis. So they set a date for a release version and hit their date quite successfully, I would say. The game is not perfect, but software is rarely in a state of perfection when planning to a fixed release date. It is not fair to the programmers to expect them to synchronize their development perfectly with the plans of the salesmen. Programming at this level simply does not work that way.
The second, and, I think, more important reason that we're essentially functioning as "beta testers" for CiV is that
this is the most efficient model for beta testing. Players are excellent beta testers, and in a game of CiV's size and complexity, it makes a lot more sense to let the players provide the "high level" beta testing en masse after initial release rather than employ the fleet of QA guys you'd need to work out every little bug in the system (because in a game like CiV, "late beta" testing requires more quantity of testing than quality). With the game handled over Steam, the update process is also greatly streamlined, meaning that users can report a bug one day and get a patch for it within a week if it's problematic enough. We have already seen this. End-user level (read: free) beta testing is actually an excellent business model and one that is not used only in the gaming community.
I say all of this as a developer, although one who hasn't worked for a games company in the last decade, so take my opinion for what it's worth. Is it a little bit sneaky to sell a "high-beta" game and then collect free testing from customers? Sure, but it is also quite sensible, and I can't say that I mind because CiV is a truly excellent game which will only get better as the kinks are worked out.