Sorry, but I disagree. The whole lesson of eventually finding AF447 was the "Bayesian" theory.
Each piece of information is evaluated, given a probability and move on from there. As with AF447, the "LKP" is often accurate, even if other info suggests otherwise. The Inmarsat info would have been evaluated, given a probability, but not straight away, enough weight to call off the other search areas.
There is also some practicality. The assets available cannot all be switched to/from areas at will.
I think it is also clear the search teams / location have been a day or 2 ahead of what we are told the "latest" info is i.e. as info comes in, someone is evaluating and actioning it. Just not the chap in front of the cameras
Edit: reading a bit more on it, you concentrate on where you are most likely to find it. That is not the same as where you think it most likely is! The original search was in shallow water, and a small area... so if it was there, the P of finding it is much higher than the deep / enormous South Indian Ocean. The former search has now stopped since not only is the P higher it is in the Indian Ocean (or the N), but since the South China Sea has been so extensively searched, the P of now finding it there goes down. Or something like that!
My point is the second one, why are they searching at some calculated end of the arc rather than along it. It isn't known what speed the aircraft was flying at. They've highlighted a search area for Australia that is at the end of the aircrafts fuel range, at some assumed cruise speed. All they know is that the transmissions were received somewhere in the arc (north or south) during those time periods. It could have been circling a single point and remained in the width of the arc and the width is not a line, it has a width due to the lack of precision of the sat data. Remember the arc is NOT a track of the aircraft. The entire arc should be given the same POD until proven otherwise.