Hmmmm...
it won't matter unless the big companies take it up first. Ask Joe HomeUser to configure their IP version, and you'll get that 1000-yard stare (and if you're lucky, a perfunctory "huh?!"). The ISPs have to drive it, then the businesses have to follow. Only then, when company-based denial of service occurs ("Why can't I get to microsoft.com anymore?") will people call up their ISP and request help converting to IPv6.
That is, if the ISPs want to lay their reputations on the line. After all, it'll be their asses in a sling if they say IPv6 is secure, and some ugly security hole pops up (and military secrets flow out like water down a storm drain). Unfortunately, other governments have gotten there first, and their black-hat types probably have an arsenal of flaws ready to exploit as soon as a worthwhile target arises. While I'm sure the US and UK governments have started testing IPv6, there's nothing like a full-scale beta test to find errors - just ask any Windows Vista user.
Rule #1: Never buy a version 1.0 of anything - wait for the first patch, and then make sure it works.
Rule #2: Count your fingers, count your toes, count your relatives...did you really get a good deal?
It might be possible to get a wide-scale, low-cost alternative IPv6 testbed going, if you gave a little (okay, a big) financial incentive to colleges and universities to upgrade to IPv6 (not including mission-critical stuff like admissions, finances, etc - they can stay IPv4). Students generate tons of traffic, with very little mission-critical stuff that needs to be sent. Students can go back to handing in paper copies of assignments instead of emailing them, at least temporarily. If IPv6 breaks, the students can go hang out, study, play sports, whatever...no real financial loss (and net health gains in most cases). They can also be encouraged to abuse IPv6, to see if they can break it.
How many BitTorrents does it take to get to the center of an IPv6 crash? One, two-hoo, three **crunch**. [Sorry, American commercial-related humor there.]
Seriously, the only way to test IPv6 is to test it (like learning to swim by swimming instead of reading about it). Somebody's got to do it, and if anyone needs new computers on a regular basis, it's the colleges and universities. Throw them a bone, and you'll find they'll be very receptive - as will the students once they get a taste of higher bandwidth and faster computers. Maybe then decisions can be made with confidence, and who knows, maybe the students and teachers will find a few major holes that will force the standards associations to take a step back and re-work IPv6 into IPv7 (or at least 6.01) - without the potential for making currently equal users into upper-class and lower-class users.
To the standards associations: next time, ADDRESSES ONLY!!! After that, only ONE component at a time, and make sure each one works correctly before you go to the next one! Don't make IP a spaghetti protocol, you morons! [Look up "spaghetti code" for a description].
You know, they should just scrap IPv6 and go to IPv7. Like Windows should be re-written from the ground up to run correctly, and should come pre-packaged with emulated versions of all previous MS OS's for backwards compatibility. Same concept, strip all the other stuff out of IPv6 and go for addresses only. Maybe then they'll have a snowball's chance in hell of assuring reliability.
But I'm not holding my breath.