barit1, excepting that the narrowly-defined region is ill-defined and depends on the crew to identify it. In addition, the probability of total power loss in not the usual certification consideration where independent probabilities of engine failure might be used; with ICI it is one single event capable of affecting all engines simultaneously; the failures are not mutually exclusive.
The events so far have not involved all engines. It would be interesting to understand why, which might refute the probability theory. However, 2 engines out on a 747 should be manageable, but 2 out on a 787 … …
If there is another event then the situation and circumstances might be very revealing in many (unfortunate) ways.
It is possible that the justification for the current restriction has considered that most engines recovered power and were not damaged beyond immediate use; however the sample size is very small for an ‘unanticipated’ situation.
The AD is a very fine safety line which depends on human judgement, hence the need for quality advice.