The nature of the data leads to uncertainty here. I believe the ping arcs can be established with good certainty - many have repeated the calculations - but these only establish the distance from the satellite at the times of the pings. To establish a position on the final arc requires assumptions to be made and in particular, assumptions about the position and timing of the final turn south. The movement of the priority search areas over time thus reflects changing support for the various assumptions which could be made. Sadly, the hard data does not favour one assumption over another and so the range of possible endpoints is very wide indeed. At the moment, the Independent Group clearly has the floor and the search area seems to have moved to match the assumptions which underlie their calculations.