Looks like the author posted it on his personal blog, but also on the group blog LessWrong, so the link avoided the filter.

The filter doesn't work several weeks later (or even days later AFAIK).

Also, the earlier submission didn't get a lot of comments (7) or votes, so a second run is probably OK by the guidelines.