After losing a particularly lulzsy thread out of the archive due to Brb, Compromised, I've added some sanity checking to the archival script. Before, the script expected two results: a thread, or a 404. This third result, where a page existed but it wasn't 4chan, resulted in the old HTML being overwritten for the threads still available for archival.
Now, the script checks for certain META tags on the /tg/ thread page. This ensures the page being retrieved is actually 4chan, and furthermore is actually a /tg/ page on top of that. Also, it turns out the scripts were still downloading thumbnail images even if they already existed, which was wasting everyone's bandwidth. FIXXED.