Don't know how much use this is, but I have access to some newspaper indexes dating way back, and the New York Times kept pretty good track of military losses from the early postwar years through the early 60s, and I was assembling a list of some selected losses:
A note about the methodology: the dates in the index are of when the crash stories appeared in the paper, not when they happened - this was generally a day or two beforehand. So, the date I've used for the loss date is for the day preceeding the story appearing; this is really an approximate date. And of course you can't believe everything you read in the newspapers, so you have to take the type and location data with a grain of salt as well.
Hope this is of some interest use to someone...
Cheers
Chris