I'd like to be clear on what I mean by "class-based". I mean that when you create the character (or possibly shortly thereafter), you pick a class for the character that does not subsequently change. Your class choice has an enormous impact on what your character can do, such as due to different classes having disjoint (or nearly so) skills or other abilities.
This excludes more free-form games like Champions Online that don't have a clear notion of a character class. It also excludes games like Uncharted Waters Online or Final Fantasy XIV, where you do have a character class, but can and likely do switch classes often.
On paper, I like the class-based design. I like playing through content as various classes and comparing how they did. I like being able to see how changing to a different class affects the game. And in principle, offering more classes gives the opportunity to do more of that.
The problem is that MMORPGs that have a lot of classes seem to invariably be awful games. And their awfulness doesn't seem to be caused by having a lot of classes, as they'd still be bad if the game deleted most of the classes.
There have been a lot of class-based MMORPGs that I've liked, or at least thought were decent games. The most classes that any of them had was Elsword, which had 10 when I started the game and 11 when I quit.
I've played a lot of MMORPGs with more classes than that: Anarchy Online, EverQuest II, Vanguard, Aion, Vindictus, Wakfu, Aura Kingdom, Black Desert, and possibly others that don't come to mind. Awful games, all of them, at least at the time that I played them.
Which doesn't necessarily mean that they were bad games at some other point in time. Vindictus, in particular, felt like I was picking through the charred remains of a game that used to be good before the developers burned it down for no apparent reason in particular. I don't know if it was good several years ago or not, but some comments on Reddit made it sound like there was some particular patch where the developers decided to make everything except for the endgame completely trivial, so that new players would have a long slog of doing a bunch of stupid stuff before they could get to the "real" game and find out if they like it.
Or take World of Warcraft as an example. I never liked it that much in Vanilla, but it wasn't a bad game then. And since then, it has made a lot of changes that I would regard as improvements. The problem is that scaling mobs to your level only sounds good on paper until you realize that they mean scaling mobs to far below your level so that everything will be trivial. People tell me that mythic+ dungeons aren't trivial, and I believe them. But if nearly an entire game is so trivial as to be stupid and boring, and you have to slog your way through that to get to the endgame to find out that you probably won't like it, either, but for different reasons, then the game is bad, period.
A friend told me that over the last several expansions, WoW has been built toward the people who had played it for many years and come back at each expansion. It's really not intended to be accessible to new players trying the game for the first time anymore. I read somewhere that Blizzard was considering greatly shortening the leveling experience. But that doesn't fix the problem. If almost the entire game is awful, then it doesn't matter whether you play through it quickly or slowly. The optimal solution is to not play through it at all.
What seems to happen is that games add more classes as years go by, and also add more content. That could lead to games with a lot of classes tending to be older games. Having a ton of content added over the course of a decade ought to be a good thing. But developers seem to have decided that it isn't, and so they make the old content trivial and stupid to try to get players through it faster so that they can get to the latest endgame content. That is to say, they knowingly and intentionally destroy nearly their entire game, so that new players can get to the endgame faster, even if it will surely be a miserable journey along the way. Some of the games that I listed above show evidence of that, though others didn't show any evidence of having been any good in the first place.
I think that some of it is due to monetization. A "free to play" game with a pay to win endgame wants people to get to the endgame so that they'll feel the pressure to pay. New players enjoying themselves for several months in lower level content is a problem, as that doesn't pressure them to pay or quit. Some games go so far as to let players pay money to have a character jump to the level cap or near it, threatening that you'll have to spend a lot of time trudging through some intentionally awful content if you don't pay up. But that's less an incentive to pay than it is to quit the game entirely.
When I played Elsword, I could actually see the game falling apart as I played it. Each new character was stronger than the previous, to try to get people to play it. By the time the game got to ten characters (after starting with three), the last character, Lu/Ciel, was something like twice as strong as the oldest characters. So the game went through all the old characters and revamped them to be on par with the newest one. Which is to say, they made all of the characters systematically a lot stronger than what most of the game's content was built for. In Elsword, I could at least set dungeons to 4-man difficulty and then solo them to get back something like the original challenge. A lot of games don't have that as an option.
I think that it's ultimately a case of developers not knowing how to add new content without wrecking the old content. Uncharted Waters Online is the only MMORPG that comes to mind as having clearly and unambiguously solved that problem, and it's a very weird game whose approaches can't readily be transferred to other MMORPGs. Guild Wars 1 did a pretty good job of it, at least until the mess that was the PVE-only skills of Eye of the North. Though at least you could just refuse to use those skills and refuse to group with people who did, even if that meant it was much more of a solo game.