Since the engine bay was closed during the dyno run in the video, I assume the shop didn’t measure the actual engine rpm but calculated it from the roller speed. I assume your fuel cutoff is actually at 8000 rpm or thereabouts yet it your printout shows 7735. The dyno can assume the wrong rpm if your rear tires were slipping on the rollers, were worn a bit, weren’t manufactured with an outer diameter exactly as expected, the shop didn’t enter your non-OEM rear tire size correctly, they entered the wrong gearing because although your NSX was first registered in 1997 it was manufactured in 1994/5, etc. Horsepower is calculated by multiplying the measured torque with the engine speed. If the shop didn’t measure the exact engine speed, their horsepower “calculation” is based on a guess.
Regarding the transmission losses, yes, they do seem very high but they vary massively depending on which dyno you put the car on. Austrian Type-R and I both put our NSXs on Bosch dynos of the same type but in different locations. His NSX showed 60% more transmission losses than mine. WTF??? ProCar Specials put his NSX on a Maha dyno (see
here and
here) and had 160% more transmission losses than I did. On a different Maha dyno (see
here), m666 edd had 220% more transmission losses. I really don’t know why the transmission losses should vary so massively except for measurement error but from the limited number of NSX dyno charts I have seen, Maha dynos seem to show the highest losses.