One of the main problems that I've been addressing in the GEM is the basic problem of "Too much area to expand + lack of immediate rivals = Runaway 1st place". In DRJ's version, civilizations such as Inca, Mali, Russia, China, Zulu, and Australia will rocket ahead of all other civs. I don't know why, but when civs start close to each other they will often bankrupt themselves to payroll a painfully large army. In all my GEM games, Persia & Arabia never founded another city, rarely will Greece found another city.
In my own map, which I admit is nowhere near perfect, I've gotten it to Mali being the only real runaway 1st place civ, and I'll be adding a Ghana, Kanem, or Wolof Civ to keep him down (as soon as I can find another african LH).
However, I have no real way, through map manipulation, to improve the AI choices of civs that sit & stagnate. Partially I think this is due to the Empire-Based idea of power that is central to CIV 4: grow your empire big, and everything will grow with it! The problem are the models for growth: territory, population, and infrastructure. Grab land, build stuff on it. With this model, technologically advanced small civs are a rarity.
Compare this to the EU3 model from the fine folks at Paradox, where city states flourish and in some instances dominate (hansa, anyone?) because their version of growth also involves trade & diplomacy, which are mostly auxiliary in Civ 4.
Hopefully this issue will be addressed in the future.