The definition of captaincy above is quite reasonable; arguably, its application during the early years of GPWS enabled crews to discount obvious system faults. However as the system reliability improved (sensor inputs - dual / monitored rad alts or rain protected rad alt) then the use of captain’s discretion could have been reduced; I would argue (perhaps with hindsight) that it should have been removed as demonstrated by several accidents where GPWS warnings were ignored or the pull up commenced too late.
The underlying issue is that as systems improve and new technology is employed, SOPs are rarely reviewed.
EGPWS is at least ten times more reliable than the early GPWS and for those systems configured with the most accurate navigation sensor inputs (embedded GPS) the reliability may be 100 times better. Thus the installation of EGPWS probably crosses the boundary where the human is more likely to make a mistake than the ‘machine’ is, even with the safeguards of captaincy.
Whilst the TAWS events reported above (FSF European Air Safety Seminar Athens Mar 06) met the captaincy definition except they were at night, the crew could have experience increased mental workload due to recalling the rule and deciding if it applies to the situation (day/night); immediate situation assessment is high workload and unnecessary if a pull up is always flown.
Current safety thinking on threat and error management is to avoid hazards and opportunities for error i.e. the captaincy judgment on the conditions existing at the time of the warning.
In addition to the need to change SOPs as technology improves, there is also a need to change the way we think about alerting and warning equipment. GPWS alerted the crew to the threat of terrain closure; the enhanced modes of EGPWS could be considered to alert the crew to an error. Either their error or someone else’s error, but whatever the reason the aircraft is not where it should be. Thus, EGPWS is a form of active Flight Ops Quality Assurance / Flight Data Monitoring (FOQA/FDM). The use of the terrain display during those flight phases where the threat of terrain is most likely adds a further safety layer by providing the crew with the ability to detect an impending FDM event without an event being recorded.
Whatever our viewpoint of EGPWS, it is a major step forward in flight safety. We should not degrade this by allowing pilots opportunity for erroneous behavior with unnecessary captaincy.