The article does an effective thing by creating a strong interest in knowing how the author will justify himself for some of the claims he's made—and it doesn't disappoint: if you give it a chance, it gives answers to the controversies.
This is one variation of an important pedagogical concept called 'motivation'. And, ironically, after reading the article, my main takeaway was that while I appreciate the main line of thinking behind it, there seems to be a fatal weakness at the foundation where it has left out the critical role of motivation in learning.
The most controversial claim made by the article is that books don't really work. It illustrates this by bringing up issues with poor retention, and justifies it by (1) pointing out that this should be expected since books as a medium were not designed on the basis of an accurate model of human cognition/learning, but rather more by chance or expedience (2) pointing out that we know from experience that some kind of active engagement with the material (e.g. problem sets) is necessary for true learning, which books by default do not provide.
The problem with that however—and probably the main reason why people are reacting so strongly against the article—is that we know books do work, and that learning by 'transmission' does work: the form books take isn't accidental, it mirrors our communication with one another by speech. And speech works. We wouldn't be here if didn't.
And the article basically admits this much, but limits it to working for "some people sometimes". I think if you hone in on that 'sometimes,' the 'some people' isn't even necessary.
The situation may be as as simple as: 'sometimes' depends on whether there is true motivation/desire/interest to learn the material in question. (And it has to be interest in the intrinsic contents of the material: being genuinely interested in learning it because it's going to make you sound smart will not work!)
Let's say you've got some debilitating condition and a book comes out that offers instructions on how to remedy it, which rest on a body of new concepts introduced by the book. Are you going to need exercises and flash cards etc.—in short, any kind of supplement to understanding beyond the raw ideas themselves—in order effectively internalize the ideas?
I would claim no. A raw stream of ideas will be adequate as long as a demand for those ideas has been made beforehand.
So what the article is really answering is: how can we teach people things that they don't really desire to know. We may desire to "know things" in general and be motivated to read books as a consequence—but that's different from having a genuine desire to understand some specific thing, and then reading about that specific thing.
The situation with instruction in schools is largely the same thing: we want students to be 'educated' and they (sometimes) want 'education'—but that generic desire for knowledge appears to be inadequate in creating the necessary antecedent desire for knowledge on the specific subject at hand.
I see going beyond books to create "mediums in which 'reading' is the same as 'understanding'," as basically a proposal to develop surrogates for motivation/interest. And maybe for e.g. systematized childhood/teenage education that ends up being the best approach. I think it's intriguing. There are definitely cases where we don't know why we should know some things, but we should just learn them anyway.
But I'd say in the general case, you should let your genuine interest guide you, and if you don't really want to know the answers given by some book, don't read it. If you really do want to know, then issues like properly dealing with metacognition etc. while reading will fall into place naturally. Your brain knows how to update its schemas with new knowledge; it's not something we need to figure out for it.