Reverse Flight
I think this goes back to which problem is identified as primary. And that has to be low RRPM. The recommended procedure for recovery for the Robinson is rolling on throttle and lowering the collective. LTE wasn't the thing that nearly killed them.
The correlation between collective and throttle you refer to is likely to drive RRPM disproportionately lower than the given movement of the throttle, as it "overreacts" (closing the butterfly valve more than desired) at lower power settings (circa < 20") and this is exactly where you're heading even if you roll off only a little.
You may be right about the margin between horn and complete stall, but manipulating these values in training in is one thing; being behind the curve in an emergency is another. The values clearly reached a critical value on two occassions before recovery.
It will take the aircraft 2-300 feet to establish in autorotation, so the RRPM recovery from entering autorotation during this period is solely from reduction of pitch, especially if descending vertically with the wind as the aircraft appeared to be. Entering auto is not the way to get RRPM back in the first instance because of the lag as the aicraft accelerates.
The correlation is not Robinson-specific btw: The Schweizer also has the same system, as do some models of the Enstrom. It isn't even piston-specific: A similar correlation between pitch and fuel demand is used in Turbine helicopters.
Apologies if this is all a bit pedantic, but it seems to me that the focus on LTE recovery leads the discussion in the wrong direction.
TT