Having been HP for a bent SK while practicing OEI flyaway - way back in about 1985 when the QHIs hadn't worked out properly how to simulate using manual throttles - I can confirm that 20 deg ND in a SK is likely to lead rapidly to the scene of the incident. Less ND, while not necessarily changing the chance of a flyout, would have given us better attitude / VV at impact and would probably have resulted in less damage.
Also, the SK procedures and numbers as used in the RAF are designed to cope with the IFR autohover case. There is, I believe, an inbuilt assumption that large attitude changes at low level are likely to lead to (Un) CFIT, and thus the numbers are designed to get you gently out of an inherently recoverable situation without losing control ie it's not just about power margins.
Sven
Gently eradicating 33 years experience with red wine