Wouldn't you need to be careful about the balance of population and RUs on balkanised worlds? Otherwise wouldn't the imbalance from a random allocation of people and stuff require even more work to justify why one state hasn't dominated another? That is, assuming geography doesn't play a large role in it.
Fair question, as I said I haven't had a need to use this mini-system a lot but I have done two balkanised worlds with it.
MgT gives a simple mechanic for determining factions and their relative strengths under a "unified" government but for balkanised worlds I assume each faction is Notable or Significant and thus worthy of at least one major settlement.
One world came out with a faction that was indeed significantly larger in terms of population/settlements but the handful of other factions collectively prevented the big one from assuming control of the world. The big one did have effective control of the starport (by hosting the Imperial noble who oversees the world) so a quick visit might leave one thinking there was one government in control when in fact several were competing. And yeah, geography had an influence here.
The second time around I did try to balance the factions a bit by liberally spreading around the minor settlements to create a more even playing field, and added Imperial reps for each country, making starport authority more a board of directors than a single noble liaison.
Both situations allowed for great local color although quite frankly my players never really explored the political situation on either world. But I don't think either world breaks anything metagame-wise.
Regarding RUs, I don't use them IMTU as this is not the OTU. My part-time Trojan Reach campaign doesn't really use RUs either but if I need them that is what I'll use. I haven't had to detail a balkanised world for that campaign yet so I guess we'll see.
But I have used my settlement system above for Nekrino, Chalchiutlicue, Dolberg, Pryme and Cordillon - so far so good.