I imagine most of the web dev world hadn't realized that the display module of CSS3 was only in working draft status. There are some serious improvements and equally serious issues involved with this seemingly innocuous property.
I've taken the liberty of copying a letter to the whatwg and html5wg from the css3 wg. This is informative reading and well worth your time. Don't tl;dr the message.
Sep 27 (1 day ago)
to www-style, Boris, public-cssacc, whatwg, public-html, List, Daniel
Forums and lists that don't publicly enforce generally accepted rules of net etiquette tend to lose their more knowledgeable members. I, and I include myself in that class, have witnessed such collapses; they're not pretty. An example may be found at DigitalPoint's html/css forums. Except for a very few of us who can't resist the challenge of providing guidance, there are no authoritative voices left.
I got to thinking about inline-block and inline-table and whether their creating a new block formatting context meant that float descendents were enclosed. The obvious next step was to set up a test case or two. Lo and behold, it does.
I have updated my Enclosing float elements demo/tutorial to reflect these findings.
David Baron recently blogged on his impression of why xhtml2 died. It's not a pretty thing.