With Schnorr signatures and the proposed batch verification of signatures do any efforts to parallelise script verification in a single block kind of become a moot point? It’ll all come down to the choice of EC multiplication algorithm used, as discussed in the schnorr-bip.
Am I getting this right?