Belcher nonetheless was the director of the flight and crew. He thus had a duty to direct with due regard for the safety of the flight.
Yes, the examiner does have this duty and ultimate responsibility, in the event of accidents, as in this case, and as I said in my post:
IF, ..if there were a landing accident, as a result of the 'seatbelt trick', or any other distraction designed to test the applicant's awareness and ability, you are right, IF the cause was the examiner's overexposure to undue hazard.
It's a judgment call on how far to go when you are testing, or training.
Training and testing must go outside of the 'normal' parameters of what we all would call a normal safe operation.
Let's take a 'normal approach', for example. We all agree (more or less) that a good stable steady easy slide it right on in there approach is what we all shoot for.
But if you teach ONLY that, you NEVER let your student deviate. Never let him get too high or low (because that might be dangerous), that student will never learn to judge the constantly changing environment and respond in a like manner. Or in other words, never learn to fly.
He will only learn rote repetitious movements to copy which will result in a paper certification that lies to the public.
So, it comes down to the timing. When did the examiner unfasten his belt? Was it during the flare? Or on final with enough time to get the applicant's response, and still be safe? As I said in my OP, the supervisor would need the actual facts that support an actual hard violation, or was it a judgment call?
And more to the point, actually. With all things being equal, when the student says one thing, and the examiner says another, and we know the truth is somewhere in between, don't you think that safety requires that we put up with an 'eccentric' examiner rather than risk producing robot instructors?