Speculative thoughts about a gyro-stabilized a/p reversionary mode.
This is beginning to look like the X-31 scenario. Once actual airspeed is far enough from system accepted airspeed, the control system gains become set inappropriately and the control system becomes unstable.
The Captain might have had a clue from deck angle when the carts started to roll, or he may just have heard the "Cavalry Charge" through the cockpit door
I find it interesting that (if I understand things correctly) if two of three air data units malfunction, the autopilot disconnects and demands that the plane be hand-flown, just when the essential airspeed data is unreliable.
What if the autopilot in that case would revert to use inertial (gyro) data as autopilot input?
(As the quote above illustrates, a dangerous pitch-up could be eminently noticeable by anyone in the plane - except the pilots, who potentially may be so fully concentrated on deciphering the ECAM messages that they fail to notice the g or pitch increase.)
It would not be rocket science to design an inertial data driven reversionary a/p mode, which might even be automatically activated in case of an unreliable airspeed indication. Just to keep the plane on a steady pitch angle while the crew sorts out the problem.
Those modern inertial data units or "gyros" measure both turn rates and linear accelerations with high precision along all three axes, so it seems at least theoretically possible to have such a unit drive a stable 1-g flight path also without airspeed data.
I realize turbulence may lead to potentially dangerous airspeed over/underspeed excursions, as a pure inertial system would try to keep the speed over ground stable. The most obvious remedy to that which I can think of would be to descend to a slightly lower altitude using e.g. GPS as altitude source, to get more speed margin. If necessary e.g. due to weather ahead, do a 180 turn.
It seems that most (all?) serious unreliable-air-data incidents have to do with iced-up pitot tubes, which however is usually a temporary condition. This reversionary mode could help keep the plane in control until the ice melts and air data comes back.
Any thoughts? Or is this capability already existing?