Without going into how the game models forces on stick (i think it's a fixed amount for one hand only for all aircraft), i think the main issue is that the 109's elevators lock up in almost level flight above 6000km but they don't do the same at lower altitudes, even at higher IAS values.
This is what's causing me to think that maybe TAS is being factored into the "stiffness" calculations instead of IAS as it should probably be in reality.
At high altitudes TAS is much higher than IAS. With the sim's 109s you can maneuver just fine at more than 400km/h IAS on the deck, but you can't do the same up at 7000km or so. I'm not an aerodynamics expert but i've studied physics in university for a few years, so i treated it as a deductive experiment, started comparing the two situations and tried to find out what's similar and what's different in each one.
Well, if the maneuver is similar, the G pulled is about the same and IAS is the same, what else is there that changes when you go from low to high altitude? Air density, which is actually why your IAS gauge shows lower than what you're really doing (TAS).
It seems like the speed threshold for elevator stiffness is based on TAS in the sim, which is like ignoring the air density at the given altitude. This also seems contrary to some well established aviation standards. Nowadays, in the age of inertial navigation systems, GPS and automated flight management systems, it's easy as pie to know an airplane's true airspeed and even it's ground speed.
Even a cessna pilot has this much information available to him: the IAS and compass readings (which provide the indicated speed vector) are fed to the GPS unit, the GPS unit already knows the aircraft's true path from the satellite and its rate of change (the real speed over the ground), so by subtracting these two vectors the GPS can also tell you the wind direction and speed.
And yet, even today, the primary speed indicator on all aircraft, from cessnas to airliners to fighters, shows IAS even when it's easy to show TAS or GS. Why? Because IAS tells us how the aircraft
feels the air around it, which is a direct measure of how well it will respond to a given maneuver. If the air is not thick enough at high altitude and the pilots sees a GS or TAS of 400km/h, he might be tricked into pulling hard into a maneuver and stalling, but if he sees an IAS of less than 200km/h this is his indication that the aircraft will feel mushy. In a sense, IAS is not so much about how fast you travel but if the air around you is dense enough to support a given maneuver.
I have the feeling that in general, aircraft who fly at different altitudes and true airspeeds will more or less fly the same if they can maintain the same IAS. Ok, maybe a higher altitude/higher TAS scenario means more Gs needed for the same change of path, but that's just what it says: you will need to pull harder or for a longer time. It doesn't however reflect on the actual effectiveness of the controls to let you do that. Simply put, if there's enough air particles to hit a deflected control surface the aircraft turns around a certain axis at a proportional rate.
In fact, i've seen some manuals for high flying general aviation aircraft that advise the pilot to always use the autopilot above a certain altitude, because the thin air presents less resistance to control deflection and thus makes it easier for the pilot to overcontrol. So for example, if it needs a force of 5kg to move the ailerons to the stops at 5000ft, it could only need a force of 1kg to do the same at 25000ft. This balances out and in fact overcomes the reduced efficiency of the controls, to the point that autopilot use is mandatory in order not to over control the airplane: it's harder for the control to effect the same change of roll/pitch/yaw for a given deflection due to the thin air, but it's easier for the pilot to reach and surpass that deflection, again because of the thin air.
To sum up, it feels like in IL2 the 109's controls need a higher force to be deflected when flying at higher altitudes due to TAS being factored into the calculations. From reading those manuals it seems that airplane control stiffness in general is mostly dependent on air density/IAS and not TAS, actually making it easier to deflect the controls at higher altitude, but not as effective for the same amount of deflection. Of course, the big question is "how much does each effect cancel the other out, or under what conditions does one of them prevail?", i wish i knew the answer to that
If all this is true (corrections are more than welcome, i'm not claiming any expertise here, just a basic understanding of a few physics variables), the way it would be modeled in IL2 would be that when flying high our aircraft would need bigger inputs but it would be easier to reach them. Essentially, the maneuvering cap during high altitude flight would not be the needed forces on the stick but the control stops.