P-39 Expert, I am with you on respecting the tests as run at the time. But, in this case, we have one test that is markedly different from all the OTHER tests.
The basic differences between the P-39C and P-39D were armor, 2 more guns and ammunition (with an existing two relocated to the wings), and self-sealing tanks. The weight difference you keep harping on, "836 pounds," WILL NOT result in an extra 1,000 feet per minute climb difference. Period. A 12% reduction in weight does NOT give you a 37% increase in climb rate. All those differences were internal. They had no aerodynamic effect whatsoever. The only difference in the two planes was weight.
So, either the P-39C was operated at higher MAP than reported or the other P-39s were operated at lower power. There is no third option. If I'm going to say a test is recorded incorrectly, it makes a LOT more sense to me that ONE test (the P-39C) is misreported than all the rest of them (P-39D, et al) being incorrect. For all I know, the Allison in the P-39C test was "hot-rodded" by being ported and polished to give more power than normal. I DOUBT that, but there is SOMETHING wrong for sure. I hesitate to even question ONE test, but we have this wide discrepancy in climb rate that only shows up in the P-39C test. Same engine, same power below 10000'. Above 10000' the P-39C horsepower was actually slightly less.
Saying that 836 pounds accounts for it just isn't going to cut it because it doesn't account for it. The LEAST history-disturbing conclusion is that the P-39C test was run in some non-standard manner. Testing was standardized, that was the purpose of the tests, to accurately compare different models. Go to wwiiaircraftperformance and look at the Tactical Suitability Tests for the P-38, P-47 and Allison P-51. They all were deficient in climb and the recommended fix was to reduce all non-essential weight. Weight influences climb more than any other factor.