If you look at the difference in gzipped size, there is not so much difference. Is it all really worth it?
The way gzip works means that a repetitive sequence of newlines and tabs will likely be only one char in the final output.
One of the things I've thought would be interesting to try, would be if you could reduce class and id names to 1 or 2 character strings a la javascript minifiers. You'd have to do globally across all html files, and linked css and js files, so it would be a bit challenging.
...minifying XHTML documents (given that they’re actually served to clients properly, with “application/xhtml+xml”) doesn’t reduce size as much as if they were HTML.
I don't understand. Why can't the size of XHTML documents be reduced as much as HTML ones? And why would the MIME type matter?