CRISPR is pretty complicated - I'd say it's not clear whether mRNA of such length could stably produce enough enzyme in enough places. It's great when you need an antigen to be presented, but not so much when you need an entire restriction enzyme system.