Do I detect the suggestion that pilots now depend on TCAS, on EGPWS, and other 'last-ditch' systems ?
Conceptually these are error detectors, either yours or someone else’s. What has happened to the need to avoid the situations where the system may be required, the multitude of error avoiding, trapping, and mitigating strategies?
If TCAS (EGPWS) fails after all of those, then perhaps it’s just not your day – or conversely it is your day.
Reliability … it’s as good as any other aviation electronic system, but an accident … including a human failure; presumably the system has a failure alert.
“ looking out the window … so always do that when responding to an RA.”
I don’t think so. The advice I have seen is to follow the RA; use the head down instruments. This avoids possible errors from misidentification and visual illusion – mistaken altitude differences, relative bearing, etc.
Don’t invent a new procedure to cover extreme ‘hypothetical’ probabilities of failure including an encounter. It’s far more likely that you have made the error in the first instance – guard against that first of all.
“Two TCAS North of Darwin” … I think that that implied an attitude to safety, opposed to the reality of serviceability.