In regards to David Mearns comments:
In a typical search situation, you would be mapping the bottom with towed arrays. You get a wide area of coverage, with decent resolution. Hopefully, they were towing multi-beam sidescan and magnetometer.
The results are then analyzied and 'targets' of value are determined.
The next step would be to send down the AUV or the ROV to look at each target.
At this depth, it would take the ROV about 8 hours to descend to each target. The surface ship would have to stay stationary above the target when using the ROV.
With the AUV, they can program the different target locations, and the fish will fly to each and map it. As noted the descent time is about 3 hours, but once at depth, it can cover multiple sites, then surface.
I believe Mearns optimism is based on the simple fact that they were mowing the lawn, then switched to the AUV, signalling a change in search methodology. Since the AUV is not meant to cover a large search area, it would be assumed that they have found some specific targets to look at.
If you hear about them deploying the ROV, you can be assured they have located the aircraft.
About searching: Depending on the minerology, and depth of sediment, these can provide all sorts of false readings on a sidescan and/or the mag.
I see quite a bit of marine traffic crossing the search path, with a lot of shipping containers. In rough seas, these vessels drop containers all the time, another way to acquire false positives while scanning.