Sometime last year, there was a claim that the frequency offset in the final satellite logon was consistent with a rapid descent. I don't know whether that's still considered true after the more detailed analysis of that data.
But, yes, if someone was flying at that point, the wreckage could be a long way from the final arc. On the other hand, if they tried to extend the flight as far as possible, you could probably estimate the position by just extending the possible tracks between the last three arcs... still a big chunk of ocean, but not a million square kilometres.