Wait, this is an unpopular opinion?
I think this is close to the consensus in 2023.
Unpopular opinion: The vast majority of opinions expressed on unpopular opinion threads are relatively popular (maybe not the majority position, but usually within the top 2 or 3; or in case of an either-or position it rarely being 3-to-1 against or the like).
D&D cannot be all things to all gamers. It's not designed to be a generic game like GURPS where you build the world exactly the way you want.
Unpopular opinion: most generic games like GURPS (Hero, Savage World, and so on) also aren't really all things to all gamers or able to (readily) build exactly what you want. Most have significant places where they do well, and places where they struggle.
Having both ability scores and modifiers is redundant. Pick one or the other.
Skills and ability scores are redundant.
D&D would be much better if ability scores were abolished wholesale. Fold everything into a single proficiency bonus where you either have it on a given roll, or don't. Skills, saves, attack rolls; any time you pick up a d20, all you need to know is whether you're proficient and whether you have advantage or disadvantage.
My take: For games like D&D (where class and level is the primary method of saying how good character X is at role Y), ability scores made sense when they 1) gate-kept certain other options like minimums for certain classes, etc. (whether this is a good idea in a given system is another question), or 2) the primary way that you distinguish otherwise like individuals. The instant you have other ways of distinguishing Fighter Joe from Fighter Jim (feats, sub-class choices, skills, for non-D&D-alikes things like advantages and disadvantages), attributes become the least interesting quality a character has. If D&D continues to want to have Attributes (and I get the reason why, if only because people expect them out of D&D), then they should go back to the drawing board and come up with something interesting to do with them (suggestion: reduce their importance to base functionality like warrior to-hit/damage and caster spell DCs and relegate them to skills and such, so that people will actually play low-int wizards or low-str fighters or otherwise allow some real variability).
Folks should read The Elusive Shift.
I have. And knew or was at one remove from a fair number of folks in it.
Trying/hoping not to single either of you out here but rather just springboard from it (because every time someone suggests that people should read Peterson, it's in threads where 90% of the participants have) -- unpopular opinion: Folks on gaming forums should assume they are not unique in a discussion as to their nerd-cred expertise. Maybe not everyone here has read Jon Peterson, but most have. Maybe an individual in the thread hasn't had firsthand encounters with some/all of the founding historic individuals in TTRPG gaming's birth, but one is unlikely to be the only thread participant who has. In discussions about realistic premodern combat (honestly at the office and on another forum more than here), I can't count the number of time where it is something like person A with 8-12 years of traditional martial arts experience and 6-8 in HEMA is trying to tell person B with 10-12 in fencing and 6-10 in MMA and person C with 3 tours of duty and 20 years of SCA 'how combat really works.'
Vancian magic really isn't that bad.
More generally: many of the trappings of older systems or older versions of systems were really useful and really successful at cultivating very specific game experiences and generally seem outmoded only when many people are not attempting to capture the same game experience. That's true for Vancian magic, treasure=xp, alignment, or any number of other systems and mechanics.