What's the problem here? Planes being able to win in situations that they shouldn't be able to in real life?
I still remember when I first started AH - the one phrase that stuck out in my mind was "It's the pilot not the plane."
There any number of explanations for why a Spit8 would lose to e.g. a Boston III in AH
1) The Spit8 pilot isn't flying the plane to its limits.
2) The Boston III got a clean shot and got the kill before the Spit's turn advantage could even come into play.
3) The Spit8 pilot does fly his plane to the limits, but the Boston III uses superior ACM (turning in the vertical, energy management, etc) to get a few clean shots to get the kill.
There are any number of explanations for why a Boston III would lose in real life
1) The Boston pilot is not at all trained in ACM.
2) The Boston pilot is scared to even engage.
3) The Boston pilot can't sustain as many Gs as the Spit pilot because he never has to train for them.
4) Maybe the Boston had no flight harness so it could not pull most BFM maneuvers (else the pilot would be thrown out of his chair). This could apply to the rest of the flightcrew as well.
5) From what I've seen in AH, the Boston uses a ring and bead sight, making good gunnery under hard maneuvering exceeindgly difficult in real life.
So here we have 8 reasonable explanations for why AH deviates from what you expect in real life. But I invite you to look closer: if any of the 3 explanations for why the AH Boston wins were actually true in real life, the RL Boston would still win. Likewise, if any of the explanations for why the RL Boston loses were actually true in AH, then the AH Boston would also lose. Furthermore, not ONE of these explanations invalidates the flight model.
Remember, when we say "it's the pilot not the plane", you have to consider the whole brain->flight stick signal path. AH's signal path is a lot cleaner and more reliable than the one pilots had in real life.