Well, if its exact duplicates you're worried about then why not set up a unique index that contained the MD5 checksum of a post and prevent them from ever being allowed in the DB in the first place? That would be pretty simple to calculate and very fast, and pretty low memory overhead as well (OTOH I havent looking into the Everything code). If it was configured to quietly ignore the dupes I would guess it would be an easy fix.
You are not ready to use symrefs unless you already know why they are bad. -- tadmc (CLPM)