I went through a bunch of HTML pages, but most didn’t have much worth keeping. The main .warc I pulled had a ton of profiles and other data, though it looks like most of the site didn’t get scraped in time. I managed to extract 63,993 songs from the various .warc files. If you’ve got a way to reach you (Twitter or email works), I can share the torrent and instructions on how to rebuild and sift through everything.