From a quick browsing, it appears the theory is basically based on this: Obama was notably strong among young voters, defining them as ages 18-30. In 2012 this group will then be age 18-34, and a larger portion of voters in general, and therefore Obama gains thanks to this, and it should be enough to flip close states like Montana and Missouri.
It's admittedly flawed on many levels.
Though perhaps a sort of "reverse" age wave might occur, as Obama's weakest age group is the one most likely to die of old age by 2012. I think this idea has more merit than pbrower's.
Poll
