people who aren't playing the game primarily for the rush of power gaming can still want mechanically effective characters, it's just that these mechanically effective characters are often in the service of some other play goal (something like, "Well, I don't want to die, because then I'll never get resolution on if Frankie the Half-Elf finds her long-lost father!" for instance points to character effectiveness as a prerequisite for someone focused on the narrative element).
<snip>
Preferences are often revealed in the few instances where the two become mutually exclusive - say, someone who must choose between taking a feat that reflects her character's journey (it's a feat that lets you cast locate person, say) but is perhaps less effective than a different feat (I dunno, crossbow expert). If you're in a situation where you CAN'T have both, which one you take says a LOT about what kind of goal you have for your play experience.
<snip>
It's nothing like a heirarchy, but it IS a difference in play goals. 1e, with its Dungeon Crawl focus, had a design that rewarded optimization, so it's not surprising that Gygax would value it highly. The 2e DMG which was the first time I encountered this concept of "winning" the game with mechanics, meanwhile, was suspicious of "too powerful characters" who would be "missing out on a lot of fun" if all they focused on was optimizing (in the process, completely missing the point that optimizing IS ITSELF part of the fun for some players!).
I think a lot of 2nd ed's advice to both players and GMs is driven by a fear that if players are able to exert significant influence on the content and direction of the shared fiction (other than in very superificial ways, like choosing the colour of a character's hat) then the game will break and there will be no stories of epic fantasy.
I think that fear is reasonably justified in context!, because 2nd ed AD&D doesn't offer many tools or mechanics or even simple GMing techniques for allowing a player-driven but story-generating game.
But 25 or so years on, with a wealth of RPG design between then and now, we don't need to have the same concerns, I don't think.
That's not to say that 5e can't break - like all editions of D&D a lot of action resolution involves cobbling together disparate elements (stats, feats, equipment, spells, items, plus whatever can be leveraged in the fiction), and a lot of that cobbling together takes the form of numbers, and if the numbers exceed the design parameters then the system won't cope. (I can report from personal experience that this can happen in the lore/knowledge subsystem of 4e if you have a Sage of Ages PC: the +6 bonus to all knowledge skills would be better implemented as a "roll twice and take the best", I think. I've heard it can happen in 5e with AC, and some people think it is a risk with the -5/+10 feats.)
But I would have thought that 5e would support more robust mechanical engagement with the goal of finding a long-lost father than simply
not dying. Unlike 2nd ed, for instance, it has a working skill system, a personality/inspiration mechanic, and its enchantment spells (charm person and the like) aren't completely broken in non-dungeon environments.
If the best way to find the lost father is taking crossbow expert, because the game broadly resembles a Tarantino movie in its quest structure, then it seems that a player who chooses Locate Object has maybe misunderstood the GM's genre, or at least is pushing against it. Conversely, if the best way to find the lost father is to find the missing locket, then Locate Object seems like a more optimised choice to me than improving my character's shooting.
If the player chooses Locate Object, and the GM praises that (perhaps as the "non-power gaming choice"), and yet the GM
still feeds the player a Tarantino movie experience, then I think the GM needs to learn some new tricks. Or at least be up front that choosing Locate Object is no more meaningful a choice in terms of the content and direction of the shared fiction than choosing a character's hat or hair style. (And we don't make players pay feat slots for those, do we?)