Part of the problem is that we've become far too comfortable with our cozy, phone-number length IPv4 addressing. After all, 192.168.1.100 is much simpler to recognize and remember than 3eff:4960:0:1001::68.
It's also true that the vast majority of IT organizations have been fairly OK within their internal reserved IP ranges for the past decade or so. The onus of not just a massive renumbering effort, but required verification that all applications and services will function properly over IPv6, is more than daunting. It's basically a nonstarter for all but the biggest IT budgets.
Thus, the problem with IPv6 is that there's no perceptible benefit for most shops, but a mountain of effort required to get there. When IT budgets are already tight, that's just not going to happen.
But those problems may be overshadowed by the larger problem of disappearing IPv4 address space. This problem isn't as big of a deal as you might think at the moment, but these addresses are being eaten up at an alarming rate, particularly as China extends Internet service to outlying areas. And of course, there's the massive number of Internet-connected mobile devices.
If there's any hope of making a real push for IPv6 throughout the computing world, it has to happen soon. Every day we continue to be fat, dumb, and happy with our IANA private ranges and doing port address translation at the firewall is another day further ensconced in the inebriation of IPv4. That simply cannot scale.
Read more about adventures in IT in InfoWorld's Adventures in IT Channel.