I do not see that USN (1 x 20 = 3 x HMG) and Williams (1 x 20 = 4 x HMG) are in contrast.
USN regarded it's comparison for a weapon set-up as a whole (strike probability included)
Since the Navy does not record what is test definition consisted of, this is hypothetical. Doing an analysis by gun makes no sense since probability analysis shows that projectile strike quantity per time fired is a variable based on rof, accuracy, and time of engagement. Yes the 20 mm gun may be equal to three 50s, but at what variables, ROF is a known, but at what accuracy, what firing time? Changing accuracy, or time of fire, will/may invalidate the conclusion, and accuracy and fire time is a situational variable, how close, how good, etc. The solution is only good at the specified variables. The proper method would be to test individual projectiles to determine velocity deltas and target damage at varying ranges. With this empirical data, mathematical analysis can be performed to determine performance over a range, time, etc. and that will vary. Now maybe the Navy did this, but then they needed to specify at exactly what is the accuracy/range and firing time. If you read my response to Gilder, just comparing gun effectivity gives a much different answer to when you calculate in hit probability. When hit probability is factored in using 4 to 1 projectile effectivity, varying accuracy give a range of gun to gun ratio from 2.5 M2 to each 20mm cannon to 2 M2 to each 20mm cannon. Quite a bit of difference, especially from the original one to three Gilder used.
, while Williams analyzed just a part of the weapon system ( the terminal effect). So I'd reckon that in 'static' firing one second burst of the Mk2 Hispano is equal to one-second burst from 4 M2 BMGs, while in the 'dynamic' firing the odds are thinned to 1=3, still in cannon's favor.
Actually, I think he did it by projectile damage using both chemical and kinetic energy.
The Hispano Mk2 was able to make 600-700 rpm, the Mk5 700-800?
If we take 750 rps for Mk5, we need 3000 rounds for 4 cannons for one minute of 'trigger time' (7 kg/sec*), and 7200-7500 rounds for 6 HMGs (M3 from Sabre; 5,16-5,375 kg/sec). No great saving weight- or volume vise if we choose HMGs, while great deal of firepower is discarded. And we never talked about all aerial targets (not just the fighters), that a cannon-armed fighter is way better choice to deal with, along with a capability to hit enemy fighters.
*then we add explosive effect, and cannon is a true winner
I believe explosive effect is included in the data I used in calculations.
Tomo, you can probably make this comment if a plane has a 75mm firing 60 r/m. Just looking at throw weight vs. rof does nothing. If that 75mm hits, it will blast the plane to smithereens. Of course it must hit. Without calculating in the probability of hits vs. impact damage, one cannot estimate the impact of various armament configurations.