takata, there is one problem with your analysis of the SHOM chart (I assume you are referring to the chart in post 3138) and the initial northward drift of bodies and wreckage that you calculated occurred between June 1 and 5. This drift lies within an area that was repeatedly overflown and searched by the Brazilian Air Force over this period, and nothing was recovered belonging to the plane.
Here are the cumulative search grids by the Brazilian Air Force June1-5:
Note, the circle in the image above does not denote a search grid, only a radius for a subsequent search focus once bodies and wreckage were discovered on June 6.
It is no accident that the search for the FDR and CDR extends west 65 NM from the last position report, and south for only 40 NM from that position. (40 NM being the standard minimum radius for all quadrants in the search grid.)
I agree that surface winds can alter the direction and speed of objects being carried by surface ocean currents. The SIGMETS issued for this area on May 31 - June 1 describe the CBs as stationary. As I read the detailed French meteorology report, a similar pattern of CB complexes occurred daily between May 30 and June 3 over this general area.
Sur la période de quatre jours du 30 mai au 3 juin 2009, l’évolution des zones nuageuses observées est typique de ce qui est fréquemment observé dans la ZCIT, qui s’étend sur une largeur moyenne de l’ordre de 500 km