Gaseous,
You say "If you've got the lever down and your Nr is still decaying you should come back and haunt your engineer! (or mechanic if you live in USofA). Helicopters should be trimmed to maintain autorotation with the lever down and should do to the point where the rotor stalls."
Agreed!
Another concern beyond misrigging is the possibility of something getting stuck under the collective. Obviously this is something that we don't want to happen, but if you lose an engine and simply cannot get the collective far enough down due to a handheld radio or misrigged collective, turns were my suggestion.
Regarding the aerodynamics, I erred. You are right, the outermost region is the driven region, rather than the driving region. Shame on me. I should (do) know better.
If we agree that the retreating side is the likely area for stall then consider my remark about reverse flow. You are right in your point that the angle of attack at the blades will increase and the stall region will move outward. However, I believe the impact is relatively small and may be offset by blade angle changes caused by a push on cyclic. Airflow impinging the bottom of the disk on the retreating side in an auto is at some acute angle. In theory, if we're doing 70kts and descening at 1500 fpm (pulling number out of my butt here--actual numbers are irrelevant), the disk is seeing airflow coming at approximately 12deg angle from horizontal at the 9 o'clock position if the ship is level. If we pitch up a degree, this inflow angle becomes 13deg. Fair enough? At 80% NR the 75% rotor radius of the blade is moving in plane at roughly 423fps, or 25380fpm relative to the aircraft. Thus, at zero blade angle, where we have 12deg disk inflow, the angle of attack is 4.687deg. The blade is seeing 1500 fpm in the vertical direction and 18296 fpm in the in-plane direction (25380 - 7084). When the disk pitches back a degree, the angle of attack is 4.956deg, a change of only 0.269deg for the 1deg disk pitch up.
Also, consider that when you pull back on the stick you are increasing pitch on the advancing side (primarily--a bit over the nose as well owing to the delta-3) and decreasing it on the retreating side. Therefore pushing forward cyclic will lead you closer to stall on the retreating side. Whether pulling back enough to yield a one degree disk pitch up will give more or less than a 0.269deg reduction in local blade angle, I don't know.
Please realize the above contains many assumptions and I've been quite loose with theory (delta_blade angle is a function of delta_angle of attack for instance). I think it helps to bring out the factors affecting the angle of attack however. I'd still like to delve a bit deeper to feel better about this analysis.
"No one really knows what the dead pilots of all the crashed R22s did last. I suspect some of them did a big flare with low Nr because they thought it would increase Nr. Imagine their surprise when the rotor stalled."
Perhaps they pushed forward
<edited to change "กใ"s into more meaningful terms>