I do not have the skills to plot/chart the results of the tests for the P-39.
What I am referring to is that the test results for the P-39N that form the basis for this argument seem to be out of the normal expected results.
That is like testing 10 aircraft and finding 8 are within 2-3% of each other, one is 5% slow and one is 5% fast. The slow one could be checked for poor engine (multiple reasons) or poor airframe causing drag. The fast one has only a few answers, bad test instruments, out of the ordinary weather conditions or an engine that is making more than ordinary power.
Please note that in many of these tests they had no way of actually KNOWING what the engine was making for power, they simply recorded the RPM and manifold pressure and consulted a chart to get the power. Some radial engines had torque meters built into the reduction gearbox and by measuring the RPM and the torque reading they had the actual power that engine was making at the moment. Not what engine XX made on a test stand when they have engine YY in the airplane.
The P-39N test could be true/accurate. It is just that it is so much better than the P-39M & Q that weren't that different.
Is the N test accurate and the M & Q tests the result of poor engines?
We also have operational charts showing no difference in performance between the N and Q (or not enough to list) and cruise charts showing little or no difference in the needed by the high drag gunpod Q and the earlier aircraft.
I would note they hung a 90 imp gallon fuel tank under a MK V Spit (790lbs on a 6695lb plane) and got 560 less fpm at 20,000ft change in rate of climb while the P-39Q lost 430fpm at 20,000ft after adding 600lbs to to 7274lb plane. I would note that the Spitfire was fully tropicalized with that fugly air cleaner and tropical radiator and oil cooler setups.