Watched the NOVA special about this. What I had
not known was that at a given altitude, air pressure, and the lift dynamics of the nominal airliner's wing dictate a cruise speed where you literally have only 5kts or whatever leeway + or - over the ideal speed, or you'll go into a stall.
A few knots too slow, and you're not generating enough lift for that altitude/air pressure. Too fast, and pre-Mach 1/sonic shockwaves begin to form and you get airflow separation from the wing. (Which I guess was the point of delta or swing-wings? So some part of the wing still always "bites" as these shockwaves form...)
I had no idea the "sweet spot" in airspeed for a given altitude/air-pressure was so narrow, or you will stall or suffer loss of positive control. And I can forgive the pilot(s) for getting target fixation on (what he thought was) his airpseed because the pitot tube/airspeed indicator froze shut.
If I remember what the experts in the NOVA show were saying correctly, pilots are taught a standardized maneuver to survive this very situation with massive instrument failure, where a predetermined throttle setting, and specific aircraft pitch pretty much guarantees you a certain airspeed, no matter what the instruments say. However, this model of Airbus' fly-by-wire system does NOT move the throttles to give extra visual feedback to the crew. So they pitched the aircraft to try and force the airspeed they wanted, but didn't manually move the throttles. Plane stalled then spun...