Thanks Te_Kahu.
SASless - a very interesting paper very much a standard text and one that has been accepted world-wide and which now is part of the ICAO work on SMS.
Nick, as always, has provided an extremely interesting overview; examination of the failure rates for engines will show that the 1:100,000 is a mean figure with some below and some above. This figure was confirmed when the Risk Assessment was provided for justification of Exposure on the North Sea.
Returning to the referenced FAA work, you will see that that the engine failure rate comes into the categorisation of 'Probable'; further examination of the Risk Matrix in figure 3.9 will show that the application of 'Probable' to the grid with a likely consequence of 'Major' (or worse) puts the operation into the 'High Risk' category (which cannot be ignored).
Lastly Nick's treatise, quite reasonably because of the audience, puts the emphasis on the risk to a single pilot. ICAO, States and organisations, whilst still considering the individual pilot, have to take a wider view and address systematic risk.
My reading of the NZ CAR text (notwithstanding the incorrect categorisation of the Performance Classes) indicates that this is exactly what was being considered for the Wellington site.
Jim