Just a couple of points:
1. It was a UAL B747-400 that almost whacked the hill coming of San Francisco, not BA. The F/O was flying and that’s why the heavy types go back to the sim a bit more often now than they did originally under the AQP program.
2. While I agree that using rudder first, and correctly, is the most aerodynamically efficient technique for a V-anything cut, it can also be the chanciest. AA was doing some training on big fast airplanes (757/767) for nationals of an un-named state a few years ago and discovered that they had a less than 50% hack rate in getting the correct rudder during high-body angle V2 cuts. If you stick in the wrong rudder, that can constitute the opening scene of the last act. The solution, from AA or Boeing I’m not sure, was to have them promptly pitch down to 12 degrees, while eliminating adverse lateral motion, initially, with aileron. Upon arrival at the new pitch attitude, the candidate should note where the yoke is pointing and make the appropriate rudder input, and subsequent trim arrangements. While far from elegant, it is not fraught with risk as is the more dynamic response. It also, as has been pointed out in earlier postings, leads to smoother and more accurate responses as experience grows.
3. Additionally, between the laws of physics and economics we are unlikely to ever have a simulator motion system that will accurately reproduce the cues of a major thrust loss, with all of the potential attendant motion and g cues. A procedure that is simple and relatively foolproof is the way to go for the majority of aviators. Boeing and Airbus recognized that in their collaborative efforts to simplify recovery from unusual attitudes, understanding that for most, the simulator was the only training vehicle available.
4. To that end, I would like to add one more amplifying comment. Maxalt correctly promotes reducing PIO through application of the correct rudder and HOLDING IT. The holding it part is spot on, but many folks get caught up in the CORRECT part and decide to refine that input before the motion fun has settled down. Better to stick a bunch of rudder in the correct direction until the initial control issue is resolved, and then refine it. This technique, from a sim point of view, is even more important on a twin than on the 747-400, which is a relative pussycat after the 757/767. This is really far less an issue in the airplane where the motion doesn’t have a software loop to get through, but a good drill will work in the real world too. All the Vmca and related discussion is appropriate in the context of flying away after the initial excitement has been resolved, but as per the initial entry, I reckon the aileron input was only intended for initial response in a particular circumstance.
PS. 'Squeeze and freeze' is a splendid phrase thanks; I'll remember to give the credit appropriately.