Hopefully in the 8 months between phase 3 and 4, those responsible have started from scratch. If there were some false assumptions and conclusions directing the earlier searches, then hopefully these have been identified and eliminated. The best way of doing this would be to bring in some new people. Otherwise it is all too easy to keep "confirming" previous errors.
It is important to know exactly what areas were searched initially (as opposed to what had been planned to be searched), as well as the meteorological conditions that may have impacted on these searches. If there was low cloud or rain or even haze, then lots of things would have missed by relatively fast moving fixed wings at 1000-1500'. Bodies most definitely would have been - they are difficult enough to see even with good visibility from a helicopter at 500-700'. Think how small people on the ground look from a 10 or 15 level building. And that is looking at the entire person - not just that bit above the water.
The other thing to consider is how many white caps there were on the waves. If there were lots of white caps, it can be hard to see light colored objects. Even a VS. All you see are hundreds of white caps in all directions - and you try and look to see which ones persist and may be something, and which ones disappear and are just breaking water. You can't keep asking the pilot to go back and circle every possible piece of debris or else you will never complete the assigned grid.
I suspect that air searches sound most easy and reliable to those who have never spent hours staring out over lots of empty water. If you listen to the accounts of survivors at sea, a common statement is how many aircraft flew over them with out seeing them.
I can only assume that they have established what areas were searched in the first few days, and the degree of confidence that can be attached to these searches.