Engineers look at empirical results most of all. They don’t dismiss large, 300,000+ car surveys just because they’re inconvenient to your argument.
I’ve said my piece on the reliability of battery designs over the past 5 years. Hopefully battery engineers improve their reliability moving forward. I don’t think it’s going to be easy though, as the simulated models of the internals of Li-ion cells is just such a devilishly difficult problem.