I tend to skip over the maths in these papers, partly because it is beyond me (I am an engineer
)partly because it's obvious that the subject does not lend itself to precise analysis, partly because nobody but a programmed autopilot could carry out those maneuvers optimally, but the paper does illustrate the general idea which is that a turnback
is possible and indeed under the right circumstances is very easy.
Whereas "everybody" says it will kill you, etc, etc.
The problem is that an EFATO at certain airports will give you no option but to go straight into the side of a building, which given the robustness of the average GA aircraft is a prob99 death. Not to mention the bad press if it is a convent/school/etc; more so if you do it in an N-reg