As a counter-argument, I would submit that human brains are a lot harder to change than aircraft systems and procedures. So while you are correct with all your points, the fact remains that we have to put warm bodies into seats, and as pilots of today are neither treated, paid or trained like astronauts, we will always get a certain percentage of colleagues who are prone to sub-standard behavior.
It's safe to say that all attempts to assess, motivate, select and ramp-check these people out of the cockpit have failed up to now...
So, one of the more productive ways of improving the overall safety record is to improve systems, interfaces, cockpit ergonomics, procedures and checklists so that even the "worst" pilots tend not to screw it up with fatal consequences.
In order to do that, the kind of "post-mortem" analysis we have seen on this thread ultimately delivers the clues needed to improve the system.
That being said, of course the idiot-proof system does not exist, and probably never will. Of course we all need to be more vigilant towards the normalization of deviation. And we also need to be less tolerant towards colleagues who underperform.
As we were heading to Dubai a couple months back, I was remarking to my sim-partner that it had been quite a few year since I'd approached a recurrent check with trepidation, much less the fear of failing to meet the required standard. Now, I'm not that good, so it can only mean that maybe some of the hurdles we are required to jump over these days are a bit low... I'd be interested to hear some of your opinions on that.