I wonder if, once they get a rocket to the max calculated safe launches, which will obviously have some margin built-in, it would be sufficiently worthwhile to

I wonder if, once they get a rocket to the max calculated safe launches, which will obviously have some margin built-in, it would be sufficiently worthwhile to keep launching it with low-value payloads (or none at all) and actually test it to failure. My intuition leans toward no, obviously because launches are very expensive in terms of fuel and other stages, but also because there's no guarantee the test rocket would be representative, so for it to be really meaningful they would have to repeat the experiment several times. Still doesn't seem inconceivable though.