In fact, combat aircraft are not normally climbed at Vy, or even Vx.
There is a process by which the excess energy is determined at various speeds and altitudes - this is quite a time consuming but ends up with a plot of constant excess energy lines on a graph of height against speed / Mach No. The aircraft is then recommended to fly a height / speed / Mach No. profile which is basically a curve which joins the maxima on the constant energy curves.
For a manually flown aircraft, this is then simplified into a series of fixed IAS or IMN values at given height brackets. For a computer flown aircraft, the max.energy climb profile will be fed into the flight-control system, which will then ensure maximum energy is maintained all the way up the climb.
The reason for this is that, although this profile does not give best climb rate, it gives a good climb rate but ensures that at any point in the climb, the aircraft is as well placed as it possibly can be to enter combat manoeuvring.
I doubt however that the programmers of these computer games have access to this sort of data, and also that may not have been the approach taken to a piston-prop aircraft like the IL2, most of my military aircraft experience is, understandably, on jets. The only aircraft in this class I've done performance work on was an N3N-3, which is similar era to the IL2 but lower performance. In that case what we did was:-
(1) Calibrate the ASI so that I had an IAS .v. CAS plot. I prefer to use a GPS-racetrack method, but range course, box-circuit, trailing static, chaseplane or tower flyby are all valid means to do this.
(2) Perform a series of sawtooth climbs, between two fixed heights, at a range of speeds bracketing where we thought Vx and Vy would be. Note that it is important to be established on condition before the first height, and continue climbing through the second, otherwise the data is confused by control issues that are irrelevant to climb performance.
Then analysing the above, actual values of Vx and Vy could be obtained. Then at each speed we...
(3) Performed constant speed climbs to a good altitude (at-least FL80). This can be tedious, but it was a nice day on the coast, so we didn't mind that much.
(4) On the ground, reduce the data to a plot of density altitude against time.
(5) Then differentiate this plot with respect to time (you can do this either graphically or algebraically using a fit to the time-alt plot).
(6) For the best angle climb, further reduce this to climb angle against density altitude and re-plot.
So, the graphs from (5) and (6) give you pretty good quality and consistent data. Extrapolating off the best-rate climb graph, you can determine service ceiling for whichever minimum climb rate you have pre-determined will mark that, and also absolute ceiling if you want it. (Climbs to FL80 on a piston prop aircraft should adequately do this).
Hope this helps. I'm afraid that I don't know a site where you can find the theory on energy-speed-altitude plots - everything I have is in old course notes, but possibly somebody else around here does.
G