Lets take MH370 as an example.
Suppose the location was found. Isn’t there a chance the location could be ambiguous about whether a crew member was flying at the point of impact or not? For example, let’s say the location did not show a long glide had occurred from the last “ping”. That doesn’t mean a pilot wasn’t conscious. It just means, there was no human input. So, it would be ambiguous. Or, maybe the location is half way between the maximum gliding distance, and a “death spiral” impact point. What then? Location would mean nothing.
* Location can tell lots of things. Comparing it to last known flight data and know data such as fuel load. This alone could prove if human inputs were made after a point that all oxygen on the aircraft would have been depleted.
So, it’s possible, that as well as locating the point of impact, you also have to salvage the wreckage. Finding the CVR for the Lion Air Boeing 737 Max was a difficult activity, and they new exactly where that impacted, and it wasn’t in deep water.
* While it would be good to salvage the wreckage it might not need to be, to determine what happened or what did not happen. Certain parts would want to be salvaged other pieces imagery might be enough. The Lion Air 737 did a dive into mud, the MH370 CVR would have lost some velocity before hitting the ocean floor.
But, even if you pay enough to locate the impact site, and then pay to find and recover the FDR, that still may not show anything. Won’t it just show the transponder lost power? It still will not indicate if it was turned off, or an open circuit occurred (say due to an exploding O2 bottle). So, you still will not know...?
* FDR may not show anything or it might show lots - an exploding O2 bottle would send a number of inputs to the FDR (if able to get the data extracted). It could also have been turned off before the transponder. Now if you have the wreckage you could inspect the O2 bottle and include or exclude the exploding bottle theory. The CVR and FDR may or may not still contain data, but they are not required to know some simple answers.
So, we get back to the ALARP principal. Which, while usually applied for risk management prior to an event, can also be applied after the fact. Is it worth spending $X for a possible increase in safety of Y?. Since $’s are finite, they should be spent on the lowest hanging fruit. And, given the Swiss cheers model, do we really need to know exactly the cause? If there are a few possible causes, wouldn’t it be prudent to try and reduce or eliminate them all? I assume potential failures are not only eliminated retrospectively.
As a travelling member of the public, I’d much much rather pay $1 per ticket to help monitor pilot mental health, than $1 per ticket to fund looking for MH370
* The pilots do not want Big Brother in the cockpit monitoring them. Certain groups think removing the pilots from the cockpit is the best way forward, and save the $1.