Ehh, he basically just drummed up a bunch of academic studies that compare various forms of reading comprehension with paper and computer screens.
As he points out, no one has done any studies on eInk or dedicated readers. Considering how closely most dedicated devices hew to traditional paper presentation -- both in terms of physical characteristics (no backlight, eInk doesn't have a refresh rate etc) and focus (eInk readers aren't multifunction devices that encourage you to multitask), though, I'd be surprised if the cognitive process is all that much different between reading eInk and paper.
I'm not the only one who believes this, by the way: Take a look at
Sandra Aamodt's comments, among others, in an article in the NY Times awhile ago.