P-39 also caught fire after first burst
No, it did not. The only fire issues with 39s were in early YP-39 models because the airflow over the Allison in-line was insufficient to keep it cool. No deployed P-39 had that problem. All deployed P-39s had fuel tank liners (self sealing), and armor around critical components. I'm not sure about the export P-400s though.
Zero was pretty good up high too - not fast
I don't think that is correct. The problem with high altitude is that in thin air you need to go faster to maintain altitude. Any sort of maneuvering on the A6Ms part above about 22K resulted in serious loss of lift both due to loss of airspeed and changes in pitch and roll affects on airflow.
but it still enjoyed the climb advantage
At altitudes below 18,000 feet and airspeeds lower than 340 IAS that was true. With the faster Allied models the Zero had a better climb rate in standard "ground to altitude" (takeoff to ceiling tests) because both planes started slow and low. A fast moving allied a.c. in, for example, a meeting engagement, moving at 360+IAS, tended to climb better than the Zeke merely because of the superior airspeed.
That is why the Zero was WONDERFUL indeed a BRILLIANT design for low speed turning engagements, and an aluminum dog for high speed boom and zoom engagements.
Penetration or kinetic energy of the round or kinetic energy a gun shoots in a time unit isnt everything. The target, plane, is not just a plate with soft, destroyable, stuff behind it. Someone will have to dig up the study for me, but post-war, it was understood that a plane will go down to same amount of destructive energy(on average) regardless of the form - kinetic as in ability to penetrate or chemical as in incendiary and/or explosive with the shrapnels adding penetration.
Not really. One of the problems experienced by both German and Japanese interceptors was that the 20mm Hispano had poor penetration. Worse, really, than the US .50cal (so judging by some of the spec sheets above, the pen numbers on the 20mm are incorrect). The problem was that the 20mm HE round simply lacked the necessary ap cap to penetrate the armor on most allied a.c. That is why there are so very, very, very many accounts of P-39, P-40, and F4-F drivers being saved from Zeros by the mere fact of the backseat armor.
In many ways, the Zero's worst problem wasn't the low velocity HE cannon, but instead the lack of sufficient penetration. If it had mounted two 12.7mm MGs with belt loads that included some fully jacketed, ballistic tipped rounds, they'd have found all targets from B-17s down to P39s easier to destroy.
Better yet would be using explosive shells where majority of the shell's lethality(energy) would be in the explosive content and the shrapnels it would accelerate.
To kill an a.c. you need to damage a critical system. There is a lot of void space on any given WW2 a.c. Shrapnel exploding against the armored seatback of an F4-F wildcat tended mostly to punch holes in the fuselage in places that did not affect the flying characteristics of the plane. There was also, behind the seat, an inflatable rubber raft that would tend to get shot up.
An Mk 108 or Mk 103(latter with greater Ke) 30 mm Minengeschoss shell had 5-6 times the lethality of a 20mm MG151/20 Minengeschoss shell and was more likely than not to cut a single engine plane in half while the gun(Mk 108) itself was only twice as heavy.
No 30mm was going to cut a single engine plane in half unless it hit something of great structural importance. It did not have sufficient bursting charge. That said, it was much more powerful than the 20mm, and a 30mm hit on a gas tank or engine was probably usually a disaster for the target.
P-39s in USAAF service sported a 37mm cannon. It was a piece of junk as A2A weaponry goes, because it had feeding problems. But when it hit, it tended to cause massive damage. Eric Bergerud mentions a case in New Guinea where a P-39 paid a visit to a Zero and the explosions in the engine mount tore the engine of the Japanese plane.
Show me a fellow who rejects statistical analysis a priori and I'll show you a fellow who has no knowledge of statistics.
Didn't we have this conversation already?