I'm actually surprised to hear you say this Steve. Rockets are predictable. Only questioning and understanding will increase our knowledge of why things like this happen.
I would suggest looking at the pad. Was there a blast deflector, and how is the pad configured? I question sim exit velocity vs actual for many flights I've seen including my own. There needs to be more data on this.
Maybe it would be better if pad rails were longer, there was passive retention at launch, or blast deflectors were not attached to the structure of the pad?
The wind velocity at increasing altitude is interesting at Black Rock. Because it's smooth, there should only be gradual increases in wind velocity immediately above the deck, well, until you are over a thousand feet and the mountains are less of an obstacle. I know this all too well from a two stage flight last year at XPRS.
Second thing I heard in this thread is "what happens at BALLS stays at BALLS." This is funny of course, but probably really true and a little disappointing. I believe that a lot of knowledge is being concentrated with a few people and teams because of a lack of sharing. I would be interested in hearing why? Maybe there wouldn't be so many failures if information flowed a bit more freely. This is also why I believe that Class 3 project participants should be encouraged to write post-flight reports.
Do suppliers of altimeters using accelerometers with tilt protection test their products to assure they will work correctly? I can test my barometric altimeters in a vacuum chamber, but how to test an accelerometer for it's safety features? Again, is this something we are going to accept and repeat?