I guess you could find the rotational inertia of each component at the HP peak RPM then find the lower inertia caused by the lightweight stuff.

From there you could figure in work equations to find out the energy required to spin each component that fast. The heavier stuff would require more energy to spin; find the difference in the energy and that would be your power difference, albeit in joules or something that you would have to convert to HP.

That's a very generic approach as I dont have the individual inertia formulas handy. I also just woke up so I'm not sure how valid all of this is, I'll try to check back later.