Yes, this stood out to me when I read the article. The examples he gives are all pop-science, all take some thesis and making a story from it that'll be interesting to a general audience.

I feel like he's looking at this through the prism of his speciality (producing interactive learning materials), and by doing that has missed the point a little bit. He's looking a genre of books at if they're a different thing to what they actually are, it's the wrong level.

I get reading a book once doesn't translate directly to "facts retained in memory" whereas his idealised new form of media does. But as you say, his examples are edutainment. I feel like what he's suggesting only usefully applies to either textbooks (which he's separated off into a different category anyway) or to a situation where the aim is to memorise as much as possible in a single reading. Which is fine! But it doesn't really need to happen for pop science -- just getting the general gist of an idea across in an entertaining package is okay.