I'm guessing this uses Readability (http://code.google.com/p/arc90labs-readability/) engine to get the body text. If this authors submitted this, would they care to confirm or explain what they are using for this?
A more up-to-date version, including source code, available here: http://fivefilters.org/content-only/
I'm not the site owner either but just found the site to be incredibly useful.