In additional, in case of a real body, there will be a little dissipation of energy with the charging and the equivalence will be not ideal.
For an ideal system this task can be reduced to the self-evident example: just change the distance between two point charges. The work/energy balance is obvious in this case.
Denis your right, the uniqueness about a van de Graff generator is their is no electric field inside the dome. Therefore there is no electric field from the dome to oppose the addition of charges. Assuming a 100% efficient van de Graff we shall input energy of 0.5CV^2
Unless that's not the energy equation but if yes it brings up problems shown in my paper.
Max, your reasoning just contains a self-evident mistake. The hint: electric field of the charged sphere and the work done by the electrostatic field in bringing a charge.
Just study electrodynamics (electrostatics in this case) and be more humble :)