If this goes to open court it would be interesting to see how ‘the law’ interprets human behaviour.
The ‘expert’ views in this forum might overlook if it is realistic for a crew to remember every single detail of a system operation, particularly those with caveats only applicable for a limited time scale, small altitude range, and having rare exposure.
Whilst it could be argued that the altitude limit could be interpreted from audio callouts (is this applicable in all aitcraft), these calls are often used subconsciously as part of the landing skill - or only consciously checked as a how goes it if the landing task is difficult.
How capable might crews be able to relate subsystem knowledge with altitude during during high workload landing flare.
The other interaction was from the recently installed runway alerting system, where an unexpected call triggered the need to GA. Had the crew been sufficiently well trained for this situation, had the crew been exposed to the callouts, of was the particular situation sufficiently well defined by the operator who selected the programmable point of alerting.
Did this system introduce conflict, the crew were just about to touchdown, longer than normal, but sufficient for the runway length; they may have judged this as they would have done many times before - it was their norm. Yet a system call out conflicted with this assessment - like a human criticising your judgement. Was the crew’s judgement correct, do they believe their human judgement, or a rarely encountered machine call out, who's validity might not have been established previously - training or real exposure.
What was the effect of policy on using automated call outs - EGPWS always react; the runway alert was part of the EGPWS - same voice, urgency, relevance? Yet the system was marketed as ‘advisory’.
How do humans perform after a surprise, the capacity to change the course of (successful) action. What effect does startle or conflict of perception, reduce the ability of humans perform expected actions, checks, calls, using other sensory input - obvious in hindsight, but unknown at the time.
A standard of human behaviour less than that required is not been tested in court. As for the manufacturer, they could always do better, see the foreseeable and judge the ‘unforeseeable’ ( I don't believe the human could do that) in system design; with or without hindsight. No A vs B argument, but B appears to have its share of AT issues; would previous events with hidden modes of quiescent operation (Asiana) be held against them, particularly where it has been argued that improvement is warranted.
Perhaps with the industry's drift towards automation the gaps in system operation should be closed - reduce the opportunity for error. Particularly if it is judged that with the increasing complexity of systems integration, operational scenarios, and demand for efficient, yet safe operations, place too greater responsibility on pilots in surprising and rarely encountered situations.
Modern pilots may not be a able to gain levels of experience previously seen; whereas those who criticise may have, but then fail to relate this difference.
We cannot be both judge and jury; yet pilots are best placed to put themselves in the situation faced by this crew, never the same because of hindsight bias, but also because of human nature; we love to blame … but would this be a defence for the manufacturer?