Having become addicted to this series, I've been idly thinking about other conflicts that could be modeled by the system. The most challenging aspect of the design would appear to be designing the bots as I can't see how you reach a workable bot flowchart without many dozens, if not hundreds, of playtestings for just a single faction's bot, let alone what would be required for designing functional bots for multiple factions that will be in play at the same time. Are there any decent blog posts or articles on-line about how bot design for COIN works in practice?
My gut feeling would be that the base game itself would be the challenging thing to balance. If the game is balanced you can make the bots work inside that system. Of course the bots will be a bit stupid at times, so then you need to adjust by giving them some slight advantage to make up for the stupid. That model already seems "solved" in principle.
The standard model for bots is not unique or optimal, of course. There could be lot's of different bots people could come up with.
But my point is that the base game should be the focus. The bots will follow if the game is good.