If you model a variable-inertia flywheel connected to a load inertia you have fundamentally a single spinning mass with the ability to vary the mass moment of inertia. The total energy of this spinning mass is constant, so when you vary the inertia, the velocity will change in order to satisfy a constant momentum. In so doing, the energy will change. The change will exactly equal the work required to change the position of the variable position mass (flyweight). Therefore, any energy change in the load inertia will be a direct result of the work input. In other words, if you want to transfer 'X' joules from the flywheel to the load, you will need to input 'X' joules to re-position the flyweights. Is this correct?