Dynos vary, so the numbers are most useful as a relative measurement against a baseline test of the bike on the same dyno.
So, I would use this same dyno for all future tests, and use this test as your baseline for comparison.
The issue of improvement with the airbox open seems to show restriction in cfm air flow thru the filter or intake system at torque peak and higher. This creates a drop in air pressure at the throttle body entry, which impairs volumetric efficiency.
It's not terribly unusual for stock systems to show this kind of behavior.
Regarding the questions about 33% drive-line loss, these bikes usually show about 5hp driveline loss in the non-UCE 5-speeds, and I doubt that the UCE is any worse than the pre-unit models. The UCE should be a little better than the pre-unit in this regard.
So, I suspect that the crank hp rating was also optimistic.