In an ideal world the only limitation to the number of civs in an iteration would be if introducing a new one would add a unique cultural and gameplay aspect to the game. With XX number of civs I have no doubt Firaxis could come up with a Siam and a Khmer that play totally different from one another, or a Georgia and Armenia. But with a finite number of reasonable slots it becomes a question of “Khmer, Siam, and Burma” or “Khmer, Georgia, and Cree.” Cultural and regional similarities become important when one of the goals is a wide variety of available play styles and representation. SEA and the Caucasus, for instance, have very distinct cultures within themselves, but less distinct than SEA from Native America or West Africa. For that reason I think beyond the true mainstay civs, we should look at mainstay niches; I think we should always have a mainland SEA civ and a Native American civ, for instance, but not necessarily always the same one.