Based on the Perl module HTTP::Proxy, this small script converts (almost) all HTML pages into black & white, including images! This is a nice way to fool colleagues: just run the proxy and ensure they use it and they are sent back to the times of black-and-white TV. :-)
Some pages will get broken by the proxy (e.g., Wikipedia, dict.leo.org, etc.) because the filtered HTML starts with four or five random hex digits. I haven't found out why this happens, yet. There are certainly other bugs as well and probably the proxy doesn't scale well. If you have some ideas to solve the issues or other comments, please let me know ().