i think it works pretty well now.
though i am open to fine-tuning this.
for example adding a rulevar that will let air to air be subject to air stacking point rules as well. i'll look into that for the next patch.
but the reason i didnt is:
-air battles are game-wise fought on one hex but they abstract/resemble a larger area fight for dominance and thus limitations seem odd. thats why there are limitations for air->ground since that sort of combat is limited to the actual hex.
-any form of penalties due to stacking opens the game up to tricking the enemy into overstacked intercepts. (i couldnt find a nice way around this)
also the airbase penalty is for combat round 1 + 2 (its quite effective), but i'll add rulevars to extend that if player wished to combat penalties in round 3+4 as well... and to be honest i can think of no better rule do deter stacking a lot of aircraft on a single airfield appart from making that a very vulnerable strategy.... so you are forced to spread out.
From my perspective, the core cause of the mega-stack phenomenon is that the attacker can pile in as many planes from as many bases as he wants into one strike. As the defender, if you try to spread out your planes among different bases, typicaly the attacker will pick a hex that is only in intercept range of part of your force and you get killed peicemeal. Forces the defender to stack up in one hex to have any chance to perserve thier airforce.
The airfield defense penalty makes the situation even worse (IMO) because now if the defender does stack in one hex he's at a combat penalty when attacked so he get's his airforce wiped (though he can make up for it with flak protection on the hex)...and if he doesn't stack up but spreads out...he gets wiped because he gets taken down peicemeal. It's a catch-22.
Note, in this phase of the air war....the attacker doesn't care about the overstack penalty against ground targets, because his objective is not to kill ground targets, it's to wipe out the opponents fighter force...so that later he can hit ground targets without much opposition. Usualy, in this phase...I'll usualy pick the LEAST important measly ground target I can find...because it's not likely to have much flak protection. I don't want to kill ground units, I want to kill enemy interceptors...so a few turns from now when they are all gone, I can hit ground targets without worrying much.
In reality, the defending air commander could say something like "Yeah, go ahead and hit that camp of reservists typing up reports 20 miles behind the line with your 100 planes.... I'm not going to waste my 20 planes to engage to defend against that.... I'll save them for something more important." Problem is with the current mechanics, the defender doesn't have much input on whether to engage or not based on the circumstances... it's just intercept or not.
It's a tough problem to solve for sure. Some mechanisms I could think of to address it would be...
1) An extra setting for the defender (or attacker for that matter) based on agressiveness. Essentialy rather retreat percentage it would determine how much the unit was willing to stick in combat based upon the ODDS it found itself against rather then simply the damage it took (i.e. if your in a 1 vs 1 fight you wouldn't want to break off after 5% loss, if you are in a 1 vs 10 you would). Of course you might have to endure a round or two while trying to break off.
2) Giving Airbases an air-capacity (just like carriers) and imposing a readiness penalty if more air units are stacked in the base then it has capacity. This means a player would be negatively effected both on offense and on defense by basing too many planes in one hex.
3) An overstacking penalty for air to air combat....and another tag for the defender to determine if they'd intercept if doing so would make them overstacked. This would make it costly for the attacker (or defender) to pile on too many planes from multiple bases into one attack. It still could be done but they wouldn't fight as efficiently. You can rationalize this as the historical difficulties encountered when trying to coordinate the operations and timing of multiple different squadrons from multiple different bases in a single sortee.
I realize that's alot of different changes...but I can't think of any other ways to address the mega-stack air combat issue. Forcing a player to spread out over multiple airbases and instituting some uncertainty about whether a unit will intercept or not helps incentivize the defender to spread out more....but doesn't do much to deter the attacker from using mega-stacks on attack from multiple bases.... and as long as the attacker can manipulate conditions (by picking targets at the edge of intercept range) so that they'll only face part of the defenders forces in any one combat, they've got a HUGE advantage in doing mega-stack attacks. Just my opinion.