To some degree, I think the "skilled segment of players" has kind of created the feedback loop that led us here — years of passive-aggressively complaining about, memeing, and mocking players who aren't "in the loop" on optimal practices leads the broader playerbase to become self-conscious and overly-concerned about small mistakes and "parse colours", which then causes the developers to try to calm people down by just removing the ability to make mistakes. Which, of course, is definitely "easier" than trying to make human beings stop being prideful and egotistical towards their "lessers" after they become skilled at something.
———————————————————————————————————————
But that's certainly not the only factor — there's also just human nature at play.
I think that most people simply do not want to be told, "Well, you're doing it wrong, but your performance is still adequate enough to not fail completely" — which is what most of the backhanded assurances given by Discords, etc, tend to sound like to players who are trying to gauge whether they're playing correctly.
This creates a general atmosphere where, regardless of how true it is, most players tend to see the game's Jobs as having only two outcomes: "failing" (non-optimal) and "succeeding" (optimal).
And I think a lot of players just do not like the idea that they're "failing" compared to the "ideal" rotation, even if the ideal rotation is not objectively necessary just to clear content.
This is why I think it's not as simple as a lot "skill-focused" players try to make it seem — the argument, "Well, let's just make the Job harder, because you'll still be able to clear content with the easier rotation".
———————————————————————————————————————
First of all, most guides won't even explain the "easy" rotation — because most guides want to teach you the "correct" way to play, and in the FFXIV community, that's interpreted as maximizing your damage potential. The people who make guides inherently see no reason to explain how to play at a mediocre level, because they assume that people can accomplish that on their own.
So when someone goes looking for a guide to their Job, it's all or nothing — they either get exposed to things like Optimal Drift and Alternate Lines, or they get nothing at all. If someone does try to make a "consolation prize" guide, it's usually not shared very broadly, and also usually dunked-on heavily by a lot of the "skill-focused" playerbase, because... again... it's imperfect.
———————————————————————————————————————
And second of all, "good enough" is just not what most people actually want.
Players don't want to think, "I'm playing the crappy version of my rotation, because it's easy, and that's what I can handle!"
They want to think, "I'm playing my rotation correctly, and I'm doing well as a result!"
The first perspective may be more realistic, but the second perspective feels far better to someone in terms of having emotional fun while playing a game.
This is why players will gravitate towards Jobs that they feel that they can intuitively play "correctly", and become averse to Jobs that they feel like they "fail" too frequently or too easily.
Let's say there's two Jobs, "Job A" and "Job B".
Job A is easy to play optimally, but has a low damage ceiling.
Job B is hard to play optimally, but has a high damage ceiling.
Let's also say that Job B played poorly ends up doing about as much damage as Job A played optimally.
I think most players will gravitate towards Job A, by a significant margin — because they don't care about their actual raw numbers nearly as much as they care about their parse colour (which has been conditioned by the community using things like "gray" and "green" as potent insults), as well as just not feeling like they're constantly messing up and failing... because that's psychologically-unpleasant.
———————————————————————————————————————
This is something that I think the argument, "Just make it more complex — if you don't like it, you don't need to do it", fails to take into account: people would rather decisively-succeed, than struggle and fail constantly, even if the numerical outcome is identical in both cases.