Jonkster, I both agree and disagree with what you say, and it is a subject we regard as being important.
To start with the contentious point. Thank you for giving the impression of paying attention, because in truth that is probably quite helpful. Repetition plays quite a significant part of how we conduct the whole operation. For example, every flight involves the crew briefing each other on the (hopefully) salient points of that particular phase of the flight. The "take off brief" will rarely involve something the crew hasn't considered many times before, but with emphasis on current "threats and errors" what it does do is to put the salient points back into pre-frontal cortex of the brain. In essence this is the brains "post it" note and moves that particular information back into the short term memory. This is important, because the onset of a "fight or flight" response is likely to cause heavy reliance on both survival (instinctive) memory, and what has been placed into that short term memory.
Take the seat belt scenario. There are many examples of potentially survivable accidents where victims who seemingly should have been able to escape are found still strapped in their seats and then seemingly overcome by a subsequent fatal event. Interviews with survivors sometimes reveal the difficulty they had getting out of their seat. In a "panic" or survival situation where the relevant escape actions weren't preeminent in the short term memory, the survival instinct was to rely on the instinctive response. That instinctive response was often to reach for the seatbelt release where it instinctively is everyday in your car, which (of course) isn't where it is in the aircraft!
I think that in many cases of "God not this stuff again" be it ahead or behind the flight deck door, the reality is that even though you may believe you have "switched off" More often than not that information is still refreshed in the prefrontal cortex and serves the intended purpose to some degree. Even if that is not universally the case, there is only so much you can do, and it is incumbent on those responsible to do it.
Where I do agree with you, is in how the delivery provides the best level of take up so that passengers and then crew minimize the time exposed to jeopardy and therefore optimise survivability. To their credit, some airlines (air New Zealand being one notable example) have put some serious thought, effort, and money into this subject. Humour is often successful, but it can be very subjective and still becomes stale with repeated exposure. Keeping the subject fresh has a significant cost element, and beyond the regulatory requirements, cost is always going to weigh heavily.
The reality (of course) is that commercial flying is extremely safe and routine, and dealing with the necessity of rare (but unpredictable) emergencies is always going to involve compromise. To that end I think we realistically have to accept some element of compromise. However, accepting the weaknesses of human behaviour (which to a greater or lesser degree we are all guilty of) I suspect the answer lies in a combination of better technical evolution, more effective education, and better human interaction between the crew and the passengers. Unfortunately, I doubt that will be a rapid development.