It's always easy to complain about insufficient URT, though when checking the diagram figures, I don't think, their UR is that bad:
- Within 2 seconds after the AP-off alarm fired, they did have the control wheel back in the position, corresponding to the pre-AP-off position (though that position was already granny compared to what was needed).
- Within subsequent 3 seconds, they were already upside-down in a 360 roll, so, control wheel full to the other side to complete, iso trying to stop the roll.
- Another 3 seconds later, they had wings "somewhat" level.
- Catching the Nr1 still at full throttle AND correcting was 12 seconds after AP-off alarm fired. Late, though given the roll going on and bringing wings somewhat level, let's give them the benefit of the doubt. We should not forget, it's probably not a coordinated roll, so gravity and wild 3D accelerations will throw around people and stuff.
- Looking at the last 1.5 seconds, the Ground speed and Computed Airspeed largely overlap, implying a far from vertical speed-vector, or better, nearly near horizontal.
- Looking at the VS vs. PA, the VS reached nearly 0 at the end, so they nearly made it to horizontal flight, before hitting the water.
- There were 24 seconds between AP-off alarm and hitting the water......
All in all, it's killing they didn't notice the throttle position discrepancy, though the recovery wasn't that bad, I'd say. They just did have the bad luck of just running out of height.
Really shocking, that an AP-off can in just 2 seconds develop into a such an irrecoverable and catastrophic situation.
Not to say, this accident clearly proves, the AP should not "let go" when reaching the end of its control wheel authority, but "freeze" (of course, with alarm), until manually turned off, when the pilots are hands-on. THIS "AP let go" is what caused the Upset to develop (rapidly), not the pilots themselves. Shocking.
Feel free to correct me.