Sticking with the risk V reward theme, the number of PEFATO drills performed over the years compared with the apparent number of accidents would suggest that the way it is being carried out is not inherently or outrageously unsafe. The missing piece of data to conclude any assessment of risk V reward is the number of REAL engine failures at takeoff there have been. It would seem reasonable to extend the total to widen the definition to power loss at takeoff, to include birdstrike etc.. Can anyone hazard a guess as to an approximate number?
The more I reflect on this accident, and the contents of the SI, the more I am drawn to question something which isnt covered in the SI: what could have caused such an experienced pilot not abandon the manoeuvre sooner?
As well as what has been discussed re height and speed, by not flying the downwind leg parallel to the runway, he didnt have room to make the turn, even if he had had the energy; indeed trying to overcome this caused him to lose more energy. The human factors analysis doesnt address the question of whether the onset of press-on-itis was in any way caused by the fact that he was a newby Red Arrow. Were there any consequences, real or imagined from "screwing up" a piece of basic training? Had he had any difficulty with this manoeuvre in the past? Could he have repeated the manoeuvre that day, or would it have required another sortie to maintain / achieve currency? Did having an engineer in the back mean that word of a failed PEFATO would have been widespread in the squadron? These seem like very relevant human factors to me.