I think that this increases focus on the reality. The reality is what browsers actually implement. Admittedly that's different for each web site. (Some are mostly FF / Chrome, some are only IE...)
If your target audience includes a slower release cycle browser or two (my data certainly shows a lot of these in many sites), then it's easier. You probably design against the last release of the slowest changing browser that matters (to you).
I imagine that's what a lot of web content creators are doing.
If you are chasing the fast moving goal posts it gets harder.
- You need to know that a browser has changed (Chrome is harder!!)
- At worst, for each change you need to figure out if that changes your personal game
Rigorously doing that looks too hard to me. Especially in a game where I find that the available facilities still fall short of what I wanted to do ten years ago! (Keeping up is not helped by the fact that news items on the web are often produced by those I don't want to hear, ever! They swamp the useful content so much that it may be unfindable.)
Unfortunately I've noted a fair number of fast moving goalpost chasers
recently. Some of them do things I'd like to use. I'm certain that the job could be done across all browsers, but they're not doing that. I imagine that's because they're locked into technologies that have major incapabilities. The fragmentation evident in the steaming jungle of native smartphone applications is spreading into browserland, getting worse I suspect.
I haven't checked out the implications of Windows 8, but I think that the old idea of one way to design web content is teetering on the edge of a precipice. The world that's shaping up seems both more interesting and too time wasting.
At the end of the day this process has, in a way, taken us back to the situation that made us want standards in the first place.