Wow, my head hurts!
What we are really getting down to is not simply "legal" v "risk", it is the continual evolution of aviation and aviators pushing the regulatory system to catch up. Everytime the law needs calrification to prescriptive and exacting standards it is because of some percieved loophole exploited beyond the reasonable. Then, we must go through another exhaustive draft of the legislation, have thousands of arguements, then write yet another prescriptive and tiny detailed rule with which people will try and misinterpret.
And I mean try. And they will complain "gee these rules are complicated..."
As Jim L quite clearly pointed out, and I think we all agree, that the rules state a NVFR operations must operate IAW FAR 135.207:
"visual surface reference, or, at night, visual surface light reference, sufficient to safely control the helicopter."
Whilst that is clear, interpretations of the rule have been far less so, and if any further substantiation of that statement were required, look through this thread!! This rule and its loose inerpretations have been at the root systemic cause of many NVFR accidents in the US. But we need to look at what the INTENT of that rule was. Stop trying to play lawyer by interpreting a way around it, look at what it was trying to keep you safe from.
It was trying to avoid a scenario where a non instrument rated pilot found themselves without a visible horizon. 178 seconds ring a bell?
Nothing to do with autopilots - if it did then instrument skills and individual helicopter stabilities would count. They dont, it is about whether or not you have, or have the risk of, getting into a no horizon situation. Nothing to do with moon angles - if it did, then pilot age and visual acuity and humidity and dust and smoke and haze and cloud type and cloud level and even windscreen cleanliness would matter. They dont, it is about whether or not you have, or have the risk of, getting into a no horizon situation.
What it did do was allow all of those guys who did not have an instrument rating and autopilot go flying when it was perfectly safe to do so - like over LA on a clear full moonlit night - without the expense of the rating and autopilot. What do we do??? We bend it and twist it and ring it to the max so that we have to have more rules clarifying a great freedom just to keep ourselves safe. And then we complain about how many complicated rules there are!
On to NVG. According to all the FAA literature (and that of SC-196) NVG is not to be used to alter or reduce the requirements of NVFR. FAR 135.27 WAS NOT WRITTEN FOR NVG. THERE WAS NO INTENT FOR IT TO COVER NVG. ALL EXPERT ADVICE AND FAA DOCUMETS TO DATE FAIL TO GIVE YOU A LEGAL AUTHORISATION TO ALTER FAR 135.27 BY USING NVG.
All Subject Matter Expertise (SME) so far on this thread have said that NVFR over water at night using NVG in the cirumstances mentioned represent an unacceptable practice. Clearly, the operation proposed does not meet the intent of FAR 135.27, nor in my opinion, the literature supporting NVG operations and you would be both technically in breach of the law, and, probably more importantly when you go to bed at night after the accident, disregarding and ignoring SME written advice and guidance.
It is not reasonable to look to the FAA to continue to control every possible interpretation of the rules and regs particularly in the face of rapidly developing technology. At some point to protect yourself from the law, you have to act "as a reasonable person" would have acted in the same circumstance, armed with the same knowledge (or knowledge that a reasonable person would have actively obtained).
That is the line we have to get to when introducing new technology like NVG. Would a subject matter expert think that is a reasonable thing to do?