whyisthereair and ttcse, takata developed a hypothesis largely based on the wreckage sighted in the several days prior to June 6 that AF447 had deviated right and possibly attempted a return in the direction of the St. Peter and St. Paul rocks.
The Brazilian Air Force's search grids June 2 through 5 were consistent with his hypothesis. These grids were generally east of the track, and once they spotted slicks of oil and other possible wreckage from the plane east of the track, the Air Force concentrated in these areas. However, the BEA report makes no mention of the oil and debris sightings before June 6, and one can now fairly conclude these were determined to be unrelated to AF447 and are discounted.
It was a search grid to the west of the track on June 6 that led to the recovery of the first bodies and wreckage determined to have been from the plane.
A bit after takata posted his hypothesis, there were a series of posts (some by me) with various plots of the surface currents over time, showing direction and speed. These posts were deleted, presumably for straying too far off-topic. In one of the deleted posts, I included the coordinates for where the VS was recovered. I found the coordinates on another site, an unofficial one, and as I recall these were for degree, minute, and perhaps second. (It would take a bit of searching to find them again.)
The plots of the surface current consistently showed, during the first week of June, the drift to be more from west to east, and from south to north, rather than the obverse. The recoveries on the days following June 6 are increasingly north of the recoveries on June 6 and June 7, until by about June 9th or so, nothing is recovered south of the latitude of Tasil.
The inference I drew is that the impact point is almost certainly to the west of where the first bodies were recovered on June 6. How far south of the recovery latitude is conjecture without detailed and accurate data on the current.
Hindsight is always 20/20, but if the Brazilian Air Force had run search grids west of the track on June 2nd, the wreckage might have been found sooner. The grids on June 1 are along the projected track from the reported position at 0210 to Tasil; the two top grids on June 2nd are the grids to the east of the two along-the-track grids of June 1, and a search consistent with the surface current if AF447 had crashed on its track,
