Monk points out that there is a Catch-22 in 64-bit adoption: people aren't demanding it because the hardware manufacturers don't support it very well, and the hardware manufacturers don't support it very well because the people aren't demanding it.
The same could be said for IPv6. Every major operating system from Windows to OS X to the Linux and UNIX derivatives all support IPv6, and there are probably entire networks run by masochists that are taking advantage of IPv6 address space. The problem is that if I configure my own IPv6 system, and I want to go read CNN.com, I'm going to have to use IPv4 to do it.
There are ad hoc workarounds and grandiose ideas about embedding all of IPv4 inside IPv6, but the simple fact is that my sister doesn't give a damn about any of this. She dials up, she opens IE, and if the CNN.com homepage doesn't pop up after she types it into the address bar, she flips out. It's a fact: there are more users on the Internet than network administrators to educate them. As such, we are never going to be able to reliably migrate to IPv6 until there's an actual benefit to the end users to want it. Until CNN.com goes strictly IPv6, there won't be much demand to stop using IPv4 to get to it.
Of course, the public backlash of such cutting edge exclusions would probably ruin the site's advertising revenue. In practice, CNN.com -- and every other site -- is going to have to support both IPv4 and IPv6 until the lion's share of the traffic ratio tips from one protocol to the other.
As God is my witness, this will never happen.
There is no advantage to sysadmins to go through the hassle of creating new IPv6 interfaces just to placate the few superdorks who insist upon using the next gen utilities, devoid of any real advantage, for no reason other than that they can. IPv4 works. IPv4 is proven technology. IPv4 is popular. IPv6 is none of these things. Why would anyone stop using one to begin using the other?