Safeware, thanks for the reply. I understand where you're coming from wrt the combination of system (excluding humans) safety and HF (and agree, for what it's worth). I guess I was maybe mislead somewhat by your choice of example (and I'm still not sure how all this relates to post 72). After all, TAWS/GPWS type systems are really only backups to the main protection against CFIT. And that's the pilot (right?). So, it would seem odd to shift all responsbility for avoiding CFIT onto some technical gizmo - some of that responsibility surely must stay with the pilot, which would suggest it's not only legitimate but essential to take the aircrew abilities into account when determining whether the risk of CFIT is acceptable. I doubt if any TAWS could be engineered to a sufficiently high standard to provide the necessary protection to allow the aircrew to stop worrying about CFIT, as it were (I know you didn't claim that). And given that we have lots of aircraft flying around, apparently safely enough, without any kind of TAWS, it suggests that, in principle, even a relatively unreliable (1E-3) device like the one Safeware suggests would just make things safer (not necessarily true, of course, if the aircrew come to rely on it too much).
As for what you say about average vs. least level of pilot competence, I think we disagree but I'm not even sure now. And if I try and explain, I think I'll just expose my ignorance of how aircrew are trained, when they are considered to be trained, etc.
Incidentally, changing subject, what on earth does this mean (from Def Stan 00-250, Part 3, 10.5.11):
Often, the designer must combine equipment reliability predictions (or failure rates) with similar predictions for human actions informal documents such as HAZOPs, fault trees analysis, event tree analysis, probabilistic safety analyses, etc.
Is it just me (that can't parse this) or is this another case of hopelessly low quality control on what is supposed to be an authoritative MoD document (as in my post 130).