CC makes a good point regarding the unlikelihood of wake turbulence being a potential factor in the left wing drop just prior to the pullup. With the wind blowing, there is plenty of ground features in that area of the course which could produce turbulence. In fact it's not unusual to see race planes bobble a bit while rounding the pylons whether from wake or other reasons.
I'm not sure about the trim tab failure being the cause of the rolling moment because the turn did not appear to tighten. Instead, the airplane appears to roll towards a wings level attitude before the increase in gee loading becomes apparent. I've thought from the start that the pitch up was due to the trim tab failure, but have no idea how much forward stick force would have been required to overcome the trim change resulting from the loss of the tab force. I suspect it's quite allot.
I've never flown a Mustang or any such high performance piston single and the highest IAS I've ever flown at is 360 KIAS, Vmo for a Westwind I. (nearly full nose down trim) Trying to imagine the trim suddenly going to neutral or less at that speed doesn't paint a pretty picture. Even the simulated trim runaway can be a handful at high speed if you don't push the trim disconnect pretty fast. Such a sudden change of required control force at near 500 mph in a Mustang must be much worse. I tend to doubt that Jimmy or anyone else would have had much chance of stopping the pitchup. Perhaps the final report will contain an analysis of the stick force change resulting from the trim tab failure. It's my feeling that any substantial instantaneous force change would likely be unrecoverable until the speed fell way off. Hannah was luckier in that his airplane climbed long enough for him to wake up.
As for all the talk about control surface stalls and high speed stalls, I just don't think that's a player here. The way I see it with the info we have right now, the tab broke, the airplane pitched up and put Jimmy to sleep with his head forward and down. With the Gee loaded on, the airplane slows down in the climb and P-factor or lack of corrective control input (pilot unconscious) rolls the plane into what amounts to a half barrel roll ending at the VIP boxes. It never looks stalled and the flightpath could easily be explained by the scenario above. As time goes on, newly revealed facts could certainly justify altering the theory to align with the new facts.
Like everyone else, I'll be interested to see a detailed accounting of all the factual data available to the NTSB team when they release it. As much as I respect their analysis of accidents, I like to do my own as well rather than just waiting to read the probable cause finding. As of today, no public docket info is yet available.