I don’t think the paper rigorously addresses the case of infinite computing power.