The flight models of IL-2, in my opinion, are just as objective as AH. What kind of sources the developer chooses to use naturally affects the game itself, and the level of objectivity is determined by careful cross examination of existing data and picking out the 'right' from the 'wrong'.
If I remember correctly, AH relies mainly on the western data and charts(escpecially the famous NACA flight data), and uses other sources as back up material for cross examination. In the case of IL-2, Oleg chooses the TsAGI data for main material and other data for back-up.
Some of the conclusions Oleg seems to make, I admit, are quite hard to 'swallow' if you have a certain level of trust for the western data on flight performances of WWII planes. One such example is the examination on the roll rates by Soviet data and concluding Bf109 roll rates as "excellent".
But if you take one step back and stop thinking about which is 'absoultely realistic', it's quite interesting see different views on a historical basis. His observations on the 190s, LaGGs, and P-51s which were separately posted on "Oleg's Ready Room"(a room for open questioning and discussions) at the IL-2 main web page was very informative, and his techinical knowledge as a former engineer of the TsAGI give a lot of weight on his assumptions, too.
For the most part, except perhaps the roll-rates of certain planes and the way engine over-heating is modelled, personally I find the planes that are both in AH and IL-2 perform very simular in flight behavior. Especially after version 1.04 where Oleg accepted the credibility of Luftwaffe data and decided to model LW planes according to their performance charts claimed by Germany. Whether this is the result of 'succumbing to popular demand' or a purely technical decision, or perhaps the result of both factors working, is still yet to be debated since Oleg doesn't say much about his reasons behind FM changes. However, the very fact the planes react in simular ways gives credit that both AH and IL-2 arrive at a certain point where two different developers agree on how a plane should perform. This IMO proves that AH and IL-2 are both top-notch in FM modelling and adjustments.
As an AH player who enjoys the potential of IL-2(however, not the other way around I admit..), some of the planes in IL-2, in my opinion, are modelled a 'little too good' - such as E retention in La-5FNs, general acceleration of late war Yaks, but that is perfectly understandable.