The existence of a design flaw does not need any crash data to prove that it exists. It like saying that a software vulnerability needs to be proven by success

The existence of a design flaw does not need any crash data to prove that it exists. It like saying that a software vulnerability needs to be proven by successful hack attempts. The issue exists regardless. What we don't have data on is the effectiveness of recommendations designed to mitigate the risk. A repeated crash is evidence that the mitigation is not sufficient.