Recoverability from a DeepStall Condition
One of the apparent "unknowns" is whether a deep stall is recoverable once an A330 is embedded in the condition. It was never flight-trialled by Airbus (i.e. neither the possibility nor the high altitude ballistic entry conditions with various trim states and engine powers). There would be many variables including weight, CofG, pitch attitude and pilot control inputs.
One paramount determinant would be the relative airflows once stabilized in a descent at greater than 40 degrees angle of attack. The effectiveness of a tailplane is a function of its lever arm and (even more vitally) the airflows across it. The Lever arm is its distance from the wing's Centre of Pressure (which "moment arm" itself changes at 40 + degrees angle of attack,
particularly with underslung engines at high power). Wind that trimmable horizontal stabilizer into a fully nose-up deflection and its chord-wise airflows are very adversely affected. At 40 degs AoA those chordwise flows would be minimal and so the tailplane's efectiveness would be greatly compromised.
In fact the whole Lift/ weight /thrust /drag relationship becomes significantly distorted for pitch-control, once "stuck" in this aerodynamically distorted flight regime. Elevator and tailplane pitch control authority are no longer a "given" at that excessive angle of attack. The "forward flight" model has now become a "falling leaf" model.... and the wind-tunnel is now an elevator shaft. A feather on a bird works well aerodynamically. A feather falling has no aerodynamic qualities whatsoever.
Something that seems to be lacking in the BEA report is the sensory deprivation of the pilots in respect of what they'd normally expect from a stalled flight condition. Uppermost in that category would be airframe buffet. Because that would be quite absent in a deep stall (there being no bathing of the empennage in turbulent flows from the wings), the resultant eerily smooth flight would induce an atmosphere of "unreality" and nothing conducive to any perceptions (whatsoever) of stalled flight. A similar sensory deprivation occurs in a super-fast lift ascending or descending in the bowels of a skyscraper. The impression of vertical speed is quite lacking once acceleration stops. However, unlike the skyscraper analogy, in AF447 the initial downwards acceleration into the deep-stall was subliminal and undetectable. That was partly due to the inadvertently optimized pitch attitude at entry caused by an inexperienced pilot's unintended or incompetent sidestick input (as well as high power, thin air and a fully deflected THS).
To sum up, as I have said before in Pprune, the pilots were suddenly operating well beyond their experience in a very alien flight regime. It's unsurprising that they ended up non-plussed and quite out of ideas. The cutback-in of the aural stall warning amidst the cacophony of other alerts and alarms only served to confound them further. It was not conducive to any concerted (or even continued) attempts to lower the pitch attitude. In fact, it was a straight-out deterrent to doing so. That was one of the few tangibles that they had to work with, but it was lethally misleading and working 180 against the (otherwise) logical solution.
If you ever wanted to construct a labyrinthine, bizarre and Byzantine conundrum for an unprepared coterie of fat, dumb and happy long-haul cruise pilots, you'd have to go a long way beyond this pitot-ice induced nightmare of countervailing automation.