It is surprisingly difficult to make solid state memory fail. I hope that part of the investigation highlights what the 'damage' to the memory was so that the DFDR and CVR manufacturers can modify the chips to make them more resilient presumably to seawater at high pressure.
It might not have been the chip but the solder joints to the chip and no easy place to latch onto it in-situ. Still if they are finally able to read it, then it doesn't need redesigning and re-cerifying it all over again.