I have to admit chatting with Gums a bit (A fellow fighter pilot).
First, note one important fact. The F-16 is a FBW aircraft that can become locked in a deep stall (if unlucky),
and it doesn't have a T tail. The pitching moment curve shows two white marked areas that independent of stabilator position, act to trap an aircraft at high angle of attack. One curve for the upright situation, and one for the inverted situation. The moment curves are pretty symmetric about the center which includes the normal flying envelope. You should note that a stabilator is a much more effective pitch control than the elevator & THS combination typically found on airline type aircraft. (This has implications for stall recovery)
I think people are missing the significance of another highly relevant fact:
AF447 was deliberately flying with an extremely aft CG. Almost 39% MAC. The aircraft is certified to fly in that envelope, but there is an important element in that certification. The flight control system has to perform properly.
My favorite flying machine, the F-4 was allowed to approach 36% MAC assuming it was not carrying too much in the way of external hardware. If you were to disable the stability augmentation system though, that nice flying aircraft became not so nice. It wobbled around in a loose semblance of controlled flight, and this was with a CG in the lower 30% MAC range. (But not so wobbly as that recent TU-154 on Rumors & News).
Now, Gums' F-16 chart is listed for 35% MAC, also fairly aft as CG goes. The F-16 is actually unstable in the subsonic portion of its flight envelope which helps its maneuverabilty, but makes the flight control system an essential element for controlled flight. It won't fly long without it!
The F-16 pitching moment chart is probably a fair model for A330 pitch authority at extended AOA ranges. Until Airbus publishes a pitching moment curve for the A330 over an extended AOA range, we can probably assume that the F-16 chart is fairly representative of what might happen with the transport as far as shape of the curves, but not the same numbers.
The A330 is a relaxed stability aircraft. This means that the certifying authorities are allowing it to fly with a CG that ordinarily wouldn't be allowed, but is allowed because the A330 has reliable systems that make this a non-event, but in the event of a loss of the normal control systems, it has enough inherent stability (say in Mechanical Law) that it can be brought back home.
I am not however proposing that AF447 lost its flight control system
(Normal Law) until the ACARS sequence but merely noting that if something interfered with the proper functioning of the system, and if the turbulent environment that the aircraft found itself in was not conducive to continued stability, then we might have an explanation for an early loss of control at the beginning of the sequence. The aircraft has to be extremely dependent on proper functioning of its flight control system near aft CG. limits. This area could probably use more discussion.