As I recommend to others, I serve my web sites as application/xml. This is obviously not a pragmatic stance if you know the issue, but rather because I like to bleed. Nevertheless, I feel that Google could stand to do a lot better handling non-text/html MIME types.
Google does not show censored results to Chinese users, reasoning that showing unreadable results is just not useful, which I can understand. Presumably, they do not index true XHTML pages (by which I mean XHTML pages served as XML) for the same reason. I imagine they feel that the number of Internet Explorer users, whom cannot access XML pages, is a large enough majority to not bother indexing them. And I can’t really fault them for feeling that way.
However, my XML pages are accessible to IE users, using a W3C-recommended trick to fool IE. If IE accessibility is a requirement for indexing, I pass the test, and I think Google should at least index sites with this trick.
Of course, I’m sort of a web fundamentalist as my refusal to serve pages as text/html, even just to Google, shows. To further paint that picture, I will rant a little more and suggest that Google index all XHTML pages and stop covering for IE’s incompetence. Maybe the pages that don’t render will get some people to question why they are putting up with the web’s worst browser. Good grief, people. Microsoft intentionally ignores IE because they realized the web was becoming a platform more successful than their bread-and-butter operating system.