Well said.
I think a related safety issue is that the chosen methods for simulating failures in real aircraft often have operational consequences that are substantially different from the operational consequences of the usual causes of real failures.