This is a potentially ugly subject.
If you want some
fairly straightforward explanations, without too much math,
this site , and
this site have some good info.
The second one is actually a PDF file of a NACA report, a bit on the technical side.
With aircraft like the P38, one main reason it had the compressibility issues it did was because it had a relatively thick airfoil, so you are looking at blunt body compressibility. The shock waves as the Mach number locally approaches M=1. It is quite possible that unpowered you could put an aircraft into a dive, and easily reach compressibility, especially at altitude. Remember the higher altitude, the lower true velocity is for the same Mach number.
There is no way AH could model all of the affects of changing air density and atmospheric characteristics in a game like this. I done a few CFD (computational fluid dynamics) problems, on large mainframe workstations, you are looking at 8-40 hours of solid computation time to solve a steady state, compressible fluid flow around a relatively simple body. Talk about low frame rate...
Everything in the game is an approximation, it has to be.
Having said that, take a look at those web sites, and google compressibility airfoil for some good pages. Some of them are very technical, and brought back a lot of bad memories for me.
Basically, though, it boils down to comparing a simulation (and more importantly, GAME) to the real world, and yes, at altitude, it would not be at all difficult to get a relatively streamlined shape, like an aircraft to reach compressibility in a dive.
The higher you go, the easier it gets. At extreme altitudes (80k feet) where the U2 flies in the subsonic regime, the range between stall speed and Mach was less than 40 knots, IIRC.
Cheers,
Spitter, the rocket scientist.