It's a bit sad when analysis can never NOT be considered politically biased.
For the record, he's right to a degree, the likely voters seem more pro Romney, having said that if you correlate that with the strongly disapprove in the approval ratings, it's probably more of an anti-Obama vote, which we knew would be Romney's strongest appeal to the base.
It was the same in 2004, the Kerry voters seemed more enthusiastic and Kerry was even ahead at this point in that cycle (according to the RCP average).
The key issue, is that until we get close to and past the conventions, likely voter screens are useless, because a lot of people aren't engaged in the process yet. That's not me as an Obama supporter making a biased point, it's something that has been observed for a long time. We simply have no idea in July, who the likely voter will be in November. We'll have a much clearer picture of the likely voter in September/October, and if Obama is trailing then, then he definitely has something to be worried about.