Flywheels and Shooter Wheels

[Cal578]

Our robot uses a single-wheel shooter on a rotating turret. (Actually, it's two wheels close together on one axle, so they act as one wheel for purposes of this discussion.) We're considering adding a flywheel, because it is supposed to help with consistency, given the different ways balls react to the shooter (older balls get softer, so they don't shoot quite as far for the same settings). At least, that's what we've been told.

My question is, has anyone done a test to demonstrate that the flywheel helps, and by how much? How big a flywheel is needed? I enjoy trying new things, but at this point in the season (our second regional is a little over a week away) I don't want to have my team spend a lot of time and resources for something that may be of no value. Possibly it could do more harm than good (causing vibrations, increasing spin-up time, adding weight high on the robot...). But if I saw evidence that it really helps, we could be convinced to give it a go.