TowerDog "How do we fix it ... Don't have the answer ..." It would be brave person who would say otherwise.
Thoughts of survival skills are interesting; these skills are associated with a positive mindset, thinking through difficult situations, being prepared for the unexpected.
Perhaps these are the aspects which regulators should be considering, first about themselves, then safety initiatives and aircraft design, and finally about the operators.
MrSnuggles, "... the trend of misunderstanding the given information, or a lack of interpretation or imagination of what is happening ... complexity ..." Patterns can be found wherever we look (what we find is what we look for), but in this instance the mounting evidence from the small number of rare events must be taken as significant. The industry does not have the luxury of time, nor hopefully, opportunity to accumulate more data, which some safety systems rely on.
The industry should not view the human as 'lacking'; this tends to imply failure and a need to improve or mend a broken system.
In accidents like this, the human is limited by inherent capability. Momentarily, situations are encountered where there is little if any opportunity to improve human performance (like those moments of puzzlement in observing an illusion, choosing a viewpoint, but not necessarily one which everyone agrees with).
The human is not broken or in need of replacement - we do our best as we see the situation. The industry has to recognise those situations where the human could be operating close to performance boundaries, and then minimising the number of occurrences so as not to cross this boundary or by providing a larger buffer space for recovery.
The situation in this accident involved an IRU failure, which output misleading attitude, and in parallel the removal (by design) of the alerting system which identified a difference in flight deck displays; a difference which existed because of the nature of the IRU failure prevented removal of the erroneous attitude display.
Most aircraft can have this form of IRU failure and also use a comparator for alerting; however it would be surprising if all aircraft types remove the comparator in the event of an upset - replies sought please. This is an area for accident investigation and safety management.
If a particular aircraft type is an oddity, then the difference needs to be identified before exposing crews to a surprising and confusing situations.
Certification and operational regulation must focus on protecting the human from extreme conditions, and not to treat a human as a hazard to be constrained, instead, improve situations to better enable them use their unique skills of decision and adaptation.
Perhaps the regulators need a 'comparator' alert in their certification.
A side thought; the IRU is a part 25 system, the CL 600 EFIS ... did that have part 23 roots (pro-line), as used in the Challenger?