flux-resource list processing is very slow

[grondo]

As noted in flux-framework/flux-sched#1137 flux resource list is very slow at processing large amounts of resources:

With ~10K nodes and ~64 cores per node the command takes 6-7s to process a sched.resource-status payload.

Profiling shows the majority of the time is spent in resources_uniq_lines(), which currently iterates over every rank in the resources set and applies the requested format to collapse lines that are identical besides the hostlist. It should be fairly simple to come up with something that's several orders of magnitude more efficient than the current approach.