That would be useful. I guess the way I've seen this done sometimes is to use symbols like * † ‡ for content footnotes and numbers for citation endnotes? But that usually works on paper, where pages are short and you can see the content of the footnote at a glance. I'm not sure I've seen a solution for this that works on the web. It'd be nice to have an integrated Less Wrong footnote system so that we could test different ways of displaying the content. Maybe a hover-over-the-footnote-to-read-its-contents feature?