Having AI bomber missions hit random strat factories rather than fields would probably be easier to program. The bombing accuracy could be set fairly low as the idea isn't so much to influence the base war but to give a focus for player fighter activity, both for intercepting and escorting fighters. Having said that bomber players might tag onto these AI missions to get some mutual protection and to pick up free kills generated by the AI gunners' defensive fire.
Logistically this makes the most sense. Semi-scheduled, fixed missions that depart from a given field or fields and attack a specific strat target or targets (nothing says you can't string them together). Numbers and type could depend on current war status, routes can be one of a pre-planned set, and country notification automated. Given the AI mission editor is in place and the missions themselves can be pre-set with the terrain, the rest would seem to be calculations from existing environmental stats.
Having said that, there are lines of code and testing involved. It would need to be something of a priority, and that would mean there is a problem to solve. The chances of this "fixing" eny are pretty low. Seems to me eny was first introduced to help solve the imbalance "problem", and now people are viewing that solution as the problem in itself. At what point do the AI missions become the problem?
I actually like the concept a LOT; background missions that keep the arenas active whether or not people are in them (early war, anyone?). Heck, if numbers get low enough I could see AI field capture missions rolling, just so the five more or less living players have something to do, but I don't see it as a way to balance the sides. There ought to be a way to put these in the background without substantially affecting the overall results of game play.