That’s a very good point, but then we are talking about edge cases and a delay measured in weeks to months. So, the benefit from catching 1 bad drug could outweigh the benefits from releasing 100 drugs slightly earlier.

It’s not an easy choice, especially when the risks are often hypothetical and the benefits so clear.

It’s interesting looking at lists like this and considering where things broken down. https://en.m.wikipedia.org/wiki/List_of_withdrawn_drugs