I think it's good to remember that there is a species of online person call "the font nerd" (I know because I am one and have font nerd friends), and font nerds would very much like it to be important to select interesting type combinations, and many are not above rationalizing urgencies for that hobby. But I doubt anybody in the real world actually cares. Look at the site we're on!
I am often surprised by how much considered opinion non-professionals have about font choices, about readability etc.
The site we're on: I don't think we are going to agree that this is the real world... ;-)
(Ask anyone with dyslexia about typefaces and they can tell you a lot about what they like and don't like. Tell them about typefaces designed with dyslexia in mind and they may love you forever. I totally changed someone's life by introducing her to OpenDyslexic.)
I think I listed a broad enough constituency to rebut that claim. But I would agree that bad font choices (as well as bad colour contrasts) are a significant downside of all the flexibility.
"designers, brand designers, accessibility designers, book publishers" are all professional font nerds.
Teachers and "avid readers" might not be, but I dispute that either of those groups cares as much about font selection as font nerds wish they did.
Are you saying professional designers don't care about this stuff?
I'm a developer with front-end skills; I've spent a lot of time around professional designers for 27 years and I usually implement people's style guides.
(I'm also a photographer, which should disqualify me from claiming to be an artist)
And I don't mean that they have to be aware they are bound like this... many are driven subconsciously by fashion concerns even when they construct other rationalization for their preferences. Humans are intensely social and can turn nearly any kind of visible behavior or marking into a social signal.
A very small number might actually be concerned with actual usability/human factors of fonts, but most are in it for the tribal aspects of associating font usage with other "brand" or tribal identities. For the amateur producers, this can lead to cargo-cult emulation of the producers they admire. Even dissent here can fall into the same trap---consider how many times you've seen a LaTex document from a student who wants to be an author of a computer science paper, but doesn't really have anything to say (yet?)...
I like fonts. I get why people like fonts. I just don't think they matter.
Now, if Computer Modern was an important professional signal, and it was the case that Monotype owned its copyright and rapaciously charged for its use, that would matter quite a bit. But that's not the case; ironically, the one font we can think of that has professional implications is under an open license.
There is a battery of commonly-used fonts (the Microsoft fonts, we might as well call them) that are professionally important. But they're also universally available; they argue against the idea that font licensing is all that meaningful, as well.
I get what you mean -- Lobster is everywhere in the summer, and it is cute but now noticeably cheap.
But again I tend towards thinking that this is an over-egged pudding, because truly distinctive commercial fonts are as likely to have negative associations, surely? Unless you're buying something that nobody else uses, and that has the same support issues as these blogs try to scare people with.
If you want something truly distinctive, you can pay for someone to make a font suite for you, and you can consider your own needs.
The BBC have for example been transitioning their sites for years over to their "Reith" font family, which has enhanced readability and rather fewer confusable pairs. ITV (the original and largest commercial TV channel in the UK) has their own typeface, "Reem", which is rather nice work (classier than most of their TV content).
A pragmatic approach for many producers would be to pick a well-considered open-licensed sans or serif face and get a font designer to produce something distinctive (that need not be a truly complete face) to pair with it for captions, logotype, alternates etc.
that's not the problem, the problem is that there are 100x as many which are not excellent, have terrible kerning, etc. and sorting between them is really time consuming. Not to mention the category of "90% of the way there" knockoffs where you can't (because you're not an expert) tell till later that you're working with something really lame. If you're not a graphic designer, it's really difficult to navigate.
(and don't read that as an endorsement of graphic designers, hire those and you've just added a lot of form bathwater, out from which the function baby will be thrown :)