Due to inherent laziness coupled with a strong sense of self-preservation, I long-ago elected to forego willing participation in the smoking-hole version of aviation Russian roulette known as multi-engine instruction in light airplanes.
Instead, I used the MEI credentials to instruct in simulators in which VMC roll and other loss of directional control accidents have a much lower fatality rate for both pilots and IP's. Having now survived many years of button-pushing and thousands of potentially-catastrophic training events, I am even more pleased with my choice of venue, and grateful that my critical sphincter function has been spared for use in my old age.
Having also been blessed with a keen powers of observation and a deep and profound grasp of the obvious, I have concluded that flying M/E airplanes and golf have much in common. The commonality extends to both performance and training, and is centered around the premise that the player or pilot gets only one kiss at the pig without incurring a significant penalty.
IOW, the first ball hit or the first engine loss in a critical situation provides the player's only opportunity to shine. All subsequent attempts (if the player survives the first blown opportunity) come at a significant price, be it a $3.50 Pro V1 and loss of hole on a Nassau or the loss of aircraft and numerous lives in a twin.
The proof of the pudding is written in the books, both in the handicap sheets at the golf club and in the pilot records at the training center. During both initial and recurrent training, pilots are given the opportunity to practice the hard stuff (V-1 cuts and S/E missed approach) on a repeated basis until they are "back in the swing of things" and can fly to check-ride standards. IOW, it's much like hitting ball after ball on the driving range, in that the "pitch 'til you win" mentality is an accepted part of the drill and the "need to get it right the first time or die" is overlooked or ignored.
If maintained for posterity, the detailed records and video tapes of each sim session would show that even trained and experienced professional pilots fall woefully short of survival-necessary skills during the first few tries at the difficult maneuvers, even though they were deemed to be proficient only six months prior. The training center's policy of maintaining strict confidentiality of training performance is probably one of their best ideas to date, because most passengers would choose to walk cross-country rather than ride with pilots who need three tries to get it right.
In another thread, I posed the question as to the percentage of legally-current M/E pilots who could successfully execute the hard maneuvers on their first real-life attempt. Based on my observations while watching them attempt to do so, I'd put the number at less than 25%.
Y'all be careful now, y'hear?