The safest jacket?….. the one that doesn’t need to live up to it’s standard. Kind of like what you said about him, the best tests are real life conditions. Sand paper/hammer&anvil types are good for simulation but like digressing in a forum on “what should I have done different” equates to arm chair quarter backing. In controlled conditions those will perform well. Remember that the next time you have the fore sight to ask the car in front of you to stop in a way such that allows you to slide or dismount the way the jacket was tested.
Life and it’s disasters are chaotic. You can take sample tests…look at all the stats you want and get an IDEA of how things COULD go. Be almost certain that it will likely have a better chance of not applying to what reality kicks in for you.
Absolutely gear up for the wreck, but ride in a way that best limits your chances of having to rely on advertising and testing not to mention earned reputation should you have to have said gear perform to it’s standards.