In some cases that’s exactly what happens. This was a known scenario and the failure was predictable.
It would have been different if they followed all industry standards and the sub still failed - that would produce valuable data that could contribute to making our understanding of the science, as well as prevailing industry standards, better.
In cases like this you can gather all the remains and data you want, and analyse it, but only if you want to confirm what we already know - reproducing observations and confirming hypothesis is an important part of science, but everything costs money, and at some point you need to triage the studies you want to put money into.