@jcjeant - Nothing to do with any "protections" as such, travel limiters on hydraulic surfaces have been around for donkey's years (you don't want the same input:deflection ratio at cruise as you do at takeoff or on approach!). It looks to me like the limiters were functioning as they should have been - restricting surface travel while apparently in stable cruise phase and giving maximum authority when things started to go pear-shaped.
@BOAC - Indeed, and this is the similarity to the Birgenair case that I feared - in the Birgenair case it was put down to the command gradient because the Captain (ex-military jock, very senior) was PF, and the F/O (PNF) was relatively junior, despite the fact that the PNF clearly had a better handle on the situation. Why the more experienced F/O (PNF) felt he could not take command from the junior F/O (PF) I cannot fathom. In fact there was an intriguing exchange at one point where the PNF calls for "controls on the left" and starts making inputs, but shortly thereafter the PF regains control and continues handling as before (alas incorrectly). To our French cousins - would you expect "controls on the left" to be a demand for both pilots to make left bank inputs or was he in fact calling for control and overruled by his junior?
@grity (and others) - I've started noticing a reference to "squirrelly" behaviour (very sensitive controls) in Alt Law, which was earlier refuted by PJ2 - who said that the controls in Alternate were slightly more sensitive, but that the difference was barely noticeable in practice. Can we stick to what we *know*, please? The point the BEA were making was that there seemed to be insufficient hand-flying training *at altitude* in AF - the FL was the main factor, not the control law.