This is a fine take. I don't care personally, about social pillar. I don't care if I ever need to participate in it. I'm more than happy to cede the floor.
So why should I be forced to put resources towards it?
Because the game is designed with that as one of the elements of play.
Why should soccer players put effort into goal kicks, if their primary interest is the center-back position? Because that's one of the things the game is designed to feature, and being inadequate in those skills means you're trying to just
not play part of the game that you'll be expected to play.
Your 1/3/5 example, if 1 is social it's paying lip service to the idea but effectively its a 0 because gamers optimize.
Nah. 1 means you're still competent--you just don't have anything more than actual competence. You have moments where you contribute something that stands out--"flashes," as I put it--but they're uncommon because you don't have as many tools in that department.
Actually having nothing significant to contribute--
actually having NO meaningful, distinct tools--is unacceptable.
And "because gamers optimize" is just as much a white-room argument as the other stuff above. Yes,
some gamers do optimize. If you ensure that putting 1 point in something
is still a valid choice, meaning it actually IS worthwhile even if it is just mostly workhorse, then it's perfectly fine to do 1/3/5. Especially if there are flavor associations with these various picks, rather than a perfect build-it-yourself thing.
E.g.: take Exploration/Combat/Socialization. If the Wizard is 2/3/4, the Fighter may be 3/5/1. Those are in-built with the class trappings and structure, so it's no longer a simple matter of "choose the optimal thing, done." Playstyle factors into it, because different mechanical implementations can be more or less pleasing to a given player. Thematic preferences factor into it, because different story can affect whether you actually enjoy taking a particular approach. And growth factors in too--if you can spend character resources as you level to gain +1 to two different pillars, that can force diversification even if you're still trying to optimize. Say every 4 levels you get +1/+1--now a 20th level Fighter can be, at the absolute most optimal, somewhere between 8/10/1 and 5/10/4. Still specialized! But something else still grows at the same time, and the relative difference between 8/10/1 and 7/10/2 (or whatever) may be worth it.
Again, this substitutes a serious response with a referenced, but unsupported, argument. Just because people
can dump a stat does not therefore mean that doing that particular thing is identical to having nothing meaningful that you bring to the table by virtue of your class specifically.
There are tools you can give players which do not care whether they have "dumped" a stat or not. You can design tools that can make a distinctive, meaningful contribution even to things that generally don't interest the player, but which can at least let them feel it was
important, not just that they were there, but that they were there
as the particular class they chose.