That kind of implies that if all those Windows users were to turn on the relevant hooks in XP then it would all burst into life. Perhaps Vista will have it on by default and everything will just start working. (:-)
Vista does apparently enable v6 by default. If it can't get a native v6 address but it has a globally routable (ie non rfc1918, etc) v4 address it will use 6to4 and tunnel IPv6 over IPv4. If it can't get a native v4 or v6 address it will use Teredo and tunnel IPv6 over UDP over IPv4 from behind NAT to get a realworld IPv6 address. This suggests that when(if?) the majority of people are using Vista then they will have at least one real world, globally scoped IPv6 address on their machine, which applications can (if they want) use.
But we know that's not the issue - it's the routers, the DNS and all the other infrastructure stuff where it's either not available or costs a whole heap more.
For native IPv6 then yes, this is all needed. To use tunneled v6 you need to be able to find other v6 endpoints via DNS, but that's about all you need.
And I suspect that there's a large number of "Networking Professionals" who, having barely got to grips with the transition from Class A, B and C addresses to CIDR format, will find IPv6 four times as hard (or is that 2^96 times harder?).
IPv6 does do a lot of things subtly differently. People will need to be trained. If people on this list don't know what a "router advertisement" is in the world of v6, they may find things very hard going if they discover they have to roll out v6 in a hurry.
So we're back to "Who pays?" again! For training and new hardware.
Who pays for v4 training and hardware at the moment? Why won't they pay for v6 training and hardware if/when IPv6 becomes reality?