There was a 20 knts instrument error in early F6f's (found by flying with a Corsair in formation), that was later corrected, after the -3 model, I believe...and maybe retrofitted to some earlier planes. The Hellcat is a 400 mph airplane.
As for propellers, it is ideal to use an airfoil that has a good Lift/Drag ratio as well as high aspect ratio for the blades without having such a narrow chord that the low Rn (Reynolds number) adversely affects drag. Any drag equates to torque, requiring more power to turn the same RPM. The power required to turn a 12'2" prop at the same RPM as a 13'1" prop varies to the 5th exponent.
(12'2" / 13'1")^5 = 70% power required
Propeller efficiency is a function of disc area vs power. The larger the disc area for the power, the more efficient (ie, helicopter).
If 2000 hp is turning the 13'1" prop, then about 1400 hp would turn the 12'2" prop.
Maximum efficiencies at say 200 mph are 93.7% for the large prop/engine combo vs 94.9% for the smaller prop/engine. If 2000 hp is turning the smaller prop, efficiency drops to 92.8%. Real world efficiencies are about 85% of these values due to rotational and tip losses. I should point out that there is a larger difference in thrust efficiencies at lower speeds. At 100 mph, for example, maximum efficiencies are 65.2% vs 61.7% for the 13' and 12' props respectively.
Likewise, the number of blades affects the power and diameter requirements. Less blades is more efficient up to a point. Tip speeds need to be kept reasonable.
Going from 3 blades to 4 would require a reduction in diameter to 94.4% that of an equivalent 3-bladed prop (3/4)^1/5.
Where paddle blade props are concerned, more blade area is concentrated out towards the tips, where the majority of thrust comes from. Also, with a constant speed prop, the tips are almost always operating at an optimum angle of attack throughout the speed range, while the roots may stall at lower airspeeds, so it's beneficial to have more blade area towards the tip under those conditions.
[This message has been edited by wells (edited 06-13-2000).]