I'd like to also see a comparison with reading vs audiobook, (and maybe vs being read to by a live person). That would also help address their hypothesis about the feedback of the book pages.
Agreed -- my own personal experience with this sort of phenomenon is via chess.
I am a much stronger player over the board than I am online and I think the visual representation/spatial relationship is a large part of it. I just engage in a different way.
While different visual representations may seem equivalent on some level, the way our brains process them may be very, very different.
I would imagine many of us also experience a similar phenomenon via writing by hand vs. typing.
I am suddenly inspired to write our head UX guy a letter.