I'm definitely no expert here but I think the L series "hyper sensitivity to detonation" (as BRAAP puts it) theory goes something like the following. For whatever reason (poor/marginally designed/manufactured or just out dated) Nissan L series heads develop localized pockets of stagnant coolant during operation. If or when nucleic (spot) boiling occurs in those pockets, the vapor can collect and form an insulating layer between the liquid coolant and the cylinder head walls. Vapor has very little thermal conductivity (compared to water, NPG, etc.) resulting in hot localized head temperatures. These hot localized head temperatures result in a head that is prone to detonation.
It appears that Evans NPG addresses the problem by providing a higher boiling point (the temperature at which liquid turns to vapor) and, according to their literature, "any locally generated vapor is immediately condensed back to liquid coolant avoiding the development of an insulating layer of vapor." This property of the coolant keeps the head walls wet allowing for much greater localized heat transfer as compared to a vapor layer.
I don't trust marketing hype so I am looking for some real world testing. TimZ is one data point favoring Evan's NPG. If I remember correctly his setup is pushing out some serious power which would significantly tax the old L6 cooling system.
So do you think it's correct to say that a higher overall temperature in the head due to NPG's lower heat transfer is negligible assuming the cooling system has the heat rejection capacity to keep up?
Sorry ozconnection for kinda going off topic...