I think what is interesting is the different ratio of crew losses to tank losses.
The problem is (and please remember, I’m not a bitch, I’m just an Aries) that *nowhere* in any of this are there any figures showing the ratio of crew losses to tank losses. It would be more helpful if there were.
What Dupuy has collected (presumably for the very excellent research reason that “that’s what there is”, and the measures we woud like to have are not recorded) is the ratio of personnel casualties to tank casualties. Not at all the same thing. Taking the personnel strength of 2 Norfolk Yeo as 662 (by backwards calculation from number and percentage of pers cas), only 337 (about half) of these actually sit in tanks of any kind (and of these 20 are in RHQ, and 44 in Honeys from recce tp). And this is a pure armour unit; as Martin has pointed out, once you go to brigades or divisions, there will be infantry and support arms and services of all sorts from which personnel, but not tank, casualties can be counted.
To take the most extreme mismatch of the unit-level stats quoted, consider the US 86th Cavalry Recon Squadron (a battalion-sized unit). They show 426 personnel cas, and 20 tank cas. That is 21.3 personnel casualties per tank casualty. I don’t care how inflammable anyone thinks Shermans are, there is no way on earth that those personnel casualties are all tank crewmen.
All the best,