Originally Posted by A-FLOOR
The pressure variations in the turbulent air will still find their way to the IAS via the static ports, and this is of course also the reason why climbing will give you the impression aircraft is accelerating. What was also significant in the AF447 case is that the turbulence will cover up any other cues that the aircraft has stalled (stall buffet), and the FBW will mask any change in roll response of the flight controls due to the extreme AOA, as I recall that the roll rate response vs. stick deflection is linear in alternate law.
Altitude variations affect IAS via the static ports. "Turbulent air" as such does not.
Mach buffet is a violent high frequency shake of the aircraft that is very different from turbulence-induced g-variations. The airplane had left the turbulence before it stalled.
Once stalled, the airplane is effectively uncontrollable in roll, its reponse to a roll input is often opposite to that commanded, at any rate quite different from normal.