As someone that does it all the time on 10b+ record tables, not really? if you don't have extra resources to build the occasional index your DB is under-provisioned and you're close to falling over, cluster or not.
ah I forgot grandparent comment said trillion. yeah that's an order of magnitude I would distribute across a cluster if we were in active development and indexes were changing etc. if you're storing that much data you should be able to afford it :)
this can mean building some untrivial infra, so the task is much more complicated than just using cloud vm.
My point is that bottle neck and why you need cluster and not single VM is likely CPU and iops, and not data storage.