To better understand what the drift analysis shows, and why there are so significant differences between the Australian and the German prediction, I would like to know whether others have the same understanding:
The Australians analyzed the path of 95 systematically positioned (real) buoys, and picked those which passed close to Reunion to determine a likely position.
The Germans did "deploy" millions of virtual buoys close to Reunion, and calculated back where they would have been 16 months earlier.
So at the (assumed) time of crash, the Australians have real buoys in the water in an
evenly distributed pattern, and just consider those which would 16 Months later come close to Reunion.
The Germans hovever look at an
extremely uneven distribution of floating objets at the time of the crash, with only a very low number (close to zero) at the most likely location of the crash determined from other data.
So the Australians determined singe possible crash sites, while the Germans determined the distribution of likely ones? (knowing that for purely random events, likeliehood does not thelp you to predict what will happen next...)
or the flaperon was ripped off in the transonic uncontrolled dive just before impact.
Or, as already mentioned, it was ripped off by flutter after loss of damping by the hydraulic actuators (loss of the according hydraulic system due to loss of engine driven pumps) in a transsonic dive.