I think it may be time to retire the Content-Type header, putting to sleep the myth that it is in any way authoritative, and instead have well-defined content-sniffing rules for Web content.

Of course, this is after stating:

IE in particular is well known for ignoring the Content-Type header, despite this having been the source of security bugs in the past.

That right there is a good enough reason for not dumping Content-Type across the board. But let's look carefully at the other examples that Ian gives.

RSS feeds are always sniffed, regardless of their MIME type, because, to quote a Safari engineer, "none of them have the right mime type".

and

Browsers have been forced to implement heuristics to handle text/plain files as binary because video files are widely served with the wrong MIME types.

Most video file formats are relatively new and it takes time for servers to be updated so that they know about the new mime-types. This isn't a new problem, do you remember PDFs? Pointing to RSS as an example isn't really helpful either, as RSS is special, and I don't mean special in a good way, as RSS doesn't have, and will never have, an official mime-type.

The root of the problem is change. There are new mime-types today that didn't exist a year ago. In a year there will be more new ones that don't exist today. Every time a new mime-type shows up we have the same problems all over again. It's called life, get over it.

Should it be easier to configure mime-types on the server? Yes.

Should web clients be aware that new mime-types are new and servers may not be configured properly and maybe do some sniffing to ensure a good user experience? Yes.

Should web clients optionally display or inform the user when such an inconsisteny is found? Yes.

Should we throw out Content-Type? No.

Should we all go read the TAG's finding on Authoritative Metadata? Yes.