The starting point is it is known in the literature that you can do additively homomorphic encryption,
Wait a moment.
Wasn't homomorphic encryption a theoretical academic thing that has not been proved to even exist (or be possible to perform) yet ?
Correct me if I am wrong here, please.