When I got my minipro, I worked a lot with Jose, the guy behind the dyno. I was getting some unusual results and we spent weeks trying different things. It ultimately lead to him changing to the sticker on the flywheel from just a piece of electrical tape that was what he originally used. But during these tests, I manually calculated torque, power, etc using the equations you can find for calculating inertia of a flywheel. They were very close and since I was excluding the flywheel posts, knew the software was outputting the correct numbers.
As Stephen said, work is work, so if you tested the same motor under same conditions with 2 different flywheels, you should get the same result. The only problems I still have is belt slippage. I've 3d printed some parts to add bearings above and below the belt to keep it somewhat tight to the motor sprocket. I have researched buying larger sprockets for the motor side, but haven't found anything that would be a easy fix.
I've found that motors of the same brand and model, but 21.5 vs 17.5 prefer the same timing to maximize peak power.