MucMuc, you're quite right on postulating eddies and the like. The other thing to consider from FluidFlow's analysis is that it does not account for the bodies probably being underwater, perhaps at a non-negligible depth, for the first few days coming up to the surface after decay had generated enough gas to bloat the bodies.
So two things are needed, the underwater current profiles and the unknowable depth to which the bodies sank. His analysis is a good line on which to search. And with the sinkage in mind that puts the likely crash site pretty close to where the slick was found, does it not?