vilas, the complexity of modern aircraft is such that it is unrealistic to expect pilots to understand, remember, and recall all of the details, even just those deemed important. And who chooses what is relevant anyway.
Even if all pilots were to be aware of a system's limitations, it would not alter the risk of encountering an adverse situation. It is difficult, if not impossible to allocate a meaningful probability to human reliability; there are too many variables and unknowns.
One very influential safety action is to change the conditions of work so that the 'system', technical design, documentation, and procedures will help the pilot in these rare conditions.
Certification was discussed earlier in the thread. We assume that the initial certification of the 777 AT operation was judged acceptable - frequency of occurrence and consequences - in context.
Since then there may be greater exposure due to the additional focus on a bounced landing; the risk increased, but probably not re-evaluated.
The introduction of RAAS appears to further increase the exposure; furthermore assuming that the TOGA AT RAAS were to be contributory in this accident, then the outcome aspects of inappropriate TOGA selection could be re-considered.
Then what would be changed, man, machine, or environment; which would be more effective.