I am aware of the limitation to soft lead, and it should not be a problem. It is a factor in deciding what to specify in an order for a set of dies, however.
The blowups came from match shooters firsthand, and were witnessed either by the shooter or the spotter (usually the spotter). It had to be a witnessed puff at around 100 yards, not a "lost" bullet presumed to be a blowup, nor a puff at long range (which is more likely an atmospheric effect). These are loads that are "normal" long range loads, high-performance, but otherwise nothing special; VLDs shot out of .223, .243 Win, 6.5-284, things like that. There is a thread on long-range.com where I got the information. I excluded many more reports because both twist and velocity were not reported. But if you make the usual assumptions, the excluded blowups would fall with the others.
These bullets don't blow every time. Sometimes you get reports of someone having a high percentage of blowups, say 75%, with a given load on a given day, but usually you just have a single bullet blow every so often, maybe 1% on average. A lot of shooters would never detect that.
Again, these are regular handloads shot at matches with the same barrels and bullets everyone else is using: Berger, Sierra, Hornady, etc.
Eric from Berger made an interesting post on the thread about the work Berger was doing to prevent blowups. His contention was that the lead was the culprit, much as I believe. Obviously I don't know what they are doing, nor is that information available. It could well be that Berger, with their more substantial equipment, can do things like make harder cores that I cannot do with equipment I can afford.
I have no doubt that I can swage my own VLDs (ULDs). However they might only be suitable for milder cartridges, say, 6.5x47 but not full-power 6.5-284.
As for testing, I do some of that. I do things like hook accelerometers to my rifles and analyze the motion during firing. I do some finite elements programming and am working on some computational fluid dynamics (CFD) stuff. It is a hobby. The CFD is for stream research, but hopefully it can be used to make an external ballistics program someday.
Scott