/prefixlength/{length} and /prefixlength6/{length} grouped every live route
of that length by prefix, shuffled all the prefixes and joined the routes
back, so a request grew with the number of routes and listed a prefix once
for each peer that shared its newest time.
Now the first 5001 prefixes of that length are read, one index search each.
With at most 5000, the page lists 500 of them, shuffled; otherwise random
row-id lookups on the mask length index pick them, at most four per prefix
asked for, filled up from the shuffled ones. Each prefix is shown once, with
its newest route, found in one search of a new prefix and last_updated index
that replaces the prefix-only one.
Model: opus-5-5
sneak