This was already discussed in the other thread, if you consider Vref close to best glide speed, variations will deteriorate the performances. Since the AP was constantly increasing the AOA to maintain a wrong path, gliding performances were constantly reduced until reaching their minimum (stick shaker). By reaching the stick shaker they basically reached the worse gliding performances they could achieve below Vref...
I think it's almost a case of "programming" - we are most use to what is "normal" and, even with the most prepared pilot, you have to add the "factor of disbelief". Part of the programming is not to go low on the approach! The first step in solving any problem is the acquisition of the correct information. Pilots have to make their decisions in real time. This crew was faced with an extremely rare situation - you might expect one engine to fail on final approach but not both at the same time.