The more interesting point, really, is where on the efficiency spectrum it makes sense to place our institutions.
I mean, we all want libraries to continue to exist in the ebook era, right. But the old-style "library buys a single copy, and lends it to one person at a time" model, which made sense for physical books because it was a natural constraint, for ebooks is weirdly inefficient from the borrower's side and seems weirdly non-remunerative from the copyright-holder's side.
Publishers have started pushing toward pay-per-use licensing for libraries, which makes more sense for both copyright-holders (who get paid more for more popular works) and lenders (who aren't held to an artificial scarcity of a "single copy"). That makes sense as far as it goes.
But again, efficiency comes into play: If I know that my library has effectively unlimited selection (because this pay-per-use licensing makes it efficient for it to "carry" all the books licensed under this scheme) and the library is easy to use to check out books from, then there's no reason for me to buy books instead of check them out, because the hassle inherent in the old physical world is gone. And from the publisher's side, every library book loaned is basically a lost sale, so they're going to want near-retail licensing prices.
And that's where the linked essay is relevant: The current state of affairs is that we live in a strong copyright regime where the publisher's desires trump everyone else's desires. If they won't license a book to libraries for less than $10 a lend, then the library either has to pay that (which few libraries have the budget for) or not carry the book. And if you think public libraries are good, then that's a bad situation.
Which tends to point toward a situation in which copyright is no longer the best way to balance our goals of allowing broad access to culture while also recompensing authors, and to think about other ways to meet those goals.