A difficulty in resolving circular debates is having a clear and universally acceptable understanding of the problem.
Many posts choose to categorise the list of accidents above as loss of control and thus relate this to manual flight practice, training, or degrading cognitive skills. There is support for this view in
The Retention of Manual Flying Skills in the Automated Cockpit.
Similarly there are alternative categorisations with causal factors such as insufficient awareness, illusion, disorientation, or startle; these categories can also be related to training and experience depending on viewpoint. We are biased by our experiences and operations, that's normal; but these same biases may prevent agreement on solutions.
Most of the views are based on accident outcomes:
'the aircraft crashed because the crew could not recover the situation'. These overlook aspects of how the situation was encountered, the preceding events, and thus opportunities for avoiding unwanted situations.
The difficulties above are typical of seeking improvement in complex operational systems, which may resist conventional (single point of view) solutions.
The industry is being exposed to 'new' ways of approaching safety, but because these involve change they may be difficult to implement. The industry may suffer a subconscious illusion of 'we are safe enough' because of a low accident rate, an old style of safety thinking, conflict between safety and economics, and belief that safety can be 'regulated' and human performance can be improved with training, all without considering the operating environment.
If this is so, then it is necessary to resolve these issues before we are able to break out of circular debates of operational safety. The higher echelons of the industry need to adapt their safety thoughts, both on the current and evolving safety issues, and not just rely on the sharp end becoming even more adaptable in a constraining regulatory environment.
Some 'break out' reading, particularly sections 1, and pages 34 - 40, 56 - 59, section 5 (page 75),
Connecting the Dots - No Single Way.
Sorry no solutions, but if we change our thinking we might progress.