Also let's not forget the deadly effects of overpressure a bursting charge will inflict if it goes off inside the confined space of a tank, where the opening or closing of hatches suddenly becomes a large factor. Heck the effects were severe enough in a buttoned up tank hit by solid shots without any form of bursting charge.
It is my understanding that the burster charge was actually quite minor, not near enough to actually cause any kind of overpressure or whatnot. Ian Hoggs 'German Artillery' shows the 7.5cm Pzgr Patr 39 to have the explosive charge of cyclonite/wax of 16 grams or .56 ounces. The burster charge was meant to break up the AP round into several chunks to increase the shrapnel effect (spall) after a penetration and hopefully light up the ammo. But on its own I don't see 16 grams of explosive causing an overpressure effect.
Well, like I said, on the western front German tank crews observed pretty much just flame ups from struck Allied tanks, and very rarely explosions. It was apparently a very different deal on the eastern front though, where the mention of enemy tanks exploding immediately after having been hit are common.
Dmitriy Loza in his book on Red Army Shermans mentions this difference. According to him it was an ammunition issue, the T-34s fuzes would go off when the ammunition burned, thus causing the HE shells to blow up. In comparison US shells burned but the fuzes did not ignite the explosives. This saved him at one point because when his crew bailed out of their sherman, they had to hide under it due to enemy fire while it burned. It got quite hot but the ammo didn't blow up, something he felt would certainly happen on a T-34.
I think that may account for the difference.
It's an interesting debate, but you need to be careful about "instant" fire and detonations.
Take a look at British tank losses and crew casualties in Normandy. The average was 1-2 crewmen "permanently lost" for each tank destroyed, which rather ruins the idea of "instant" detonations. All the accounts I've read of crew bailing out indicates there are a few seconds between impact and destruction. The same applies to the Russian experiences I've read.
I have personally come to question this statistic recently. In the book "South Albertas: A Canadian Regiment at War", Donald Graves examine the regiments casualties. From August 1944 to May 1945, the regiment lost about 200 tanks to all causes, and suffered 100 deaths. Of those, only 10 actually died in the tanks. The rest were cut down while bailing out and/or retreating their lines on foot, killed while caught outside their tanks during a German barrage, were commanders who were killed while unbuttoned, or died due to flu and accidents. Doesnt include wounded so I imagine total casualties per tank lost were probably at least 1 per lost Sherman.
Still casualties directly related to the penetration seem to be low. Many tank casualties would be a side effect of the tank being penetrated, mainly those being cut down as they bailed out and tried to retreat back to their own lines, at least in this regiment. I am not sure who or how people came to the oft quoted 1-2 casualties per KOed tank, but I wounder if it may simply have been dividing crew casualties by tanks reported lost.