tedster - 4:05 pm on Nov 18, 2010 (gmt 0)
no evidence that browsers, screen readers or search engines treat these elements differently, so for developers to treat them differently is a fallacy.
If browser makers don't support something and coders don't follow it either, then it doesn't belong in the spec.
This whole area is a classic (but minor) case for why the WHATWG split off from the W3C. Now they're working together again and we have a new underlying philosophy for HTML5. "Pave the cow paths" is one way of saying it.
There's no reason for any change just because there is some academic or or theoretical "purity" to it. It's time to get real.