I'm expecting turnout to be higher numerically, but I think it will be lower as a percent of registered voters.
I agree. And, J. J., it doesn't depend on who turns out. Unlike 2004, it is eminently obvious who would be the ones turning out in a high-turnout election.
In the current landscape, sure, but the sheer complexity of the variables between now and November (youths deciding Obamamania was just another passing fad, the emergence of culture war along the fissure lines of gay marriage, etc etc) means no, it is not eminently obvious who benefits from a high turnout.