It is worth noting that people have been trying to find narrative options for non-spellcasters in DnD for, um, decades, without any real success.
folks were not trying to find "narrative options" for non-spellcasters for D&D for decades. Folks have been doing many things, but looking at it as "narrative options" is a pretty new concept.
The earliest example I can think of, of doing something along the lines of trying to find "narrative options" for fighters, is a Dragon article by Katharine Kerr from the mid-80s (Mar 85, Dragon 95, pp 33ff) discussing the award of XP for non-combat activity, and trying to consider how the fighter can meaningfully participate in such activity, given that (in Gygaxian AD&D, at least) full XP awards depend upon playing to your class archetype:
What really counts in awarding experience points, then, is the struggle of the PCs to succeed in whatever task the scenario sets for them. The only reason that the experience-point system as it stands is so limited [to XP for treasure recovered and monsters defeated] is that it was designed to judge a very limited type of scenario —- the dungeon adventure or its above-ground equivalent. Our problem arises when the goals of the scenario have nothing to do with armed enemies and loot, but we can expand the basic principle of using the goals of the scenario as the basis of awarding points to cover a vast number of different kinds of game actions.
Before we do, however, there is one more aspect of the rules as they stand that we must consider, namely, that the PCs must use the skills of their particular character classes in completing the adventure. . . a PC who acts outside of his class or who doesn’t use the skills specific to his class is not entitled to a full share of the points awarded by the DM. . .
Let'’s again consider the example of the PC party attempting to rescue a friend from slavery. Since one of the stipulations of the scenario is that violence is forbidden, how can the fighter in the party pursue his major aim, which is to fight? In this and similar situations, we need to extend the concept of the major aim of each class to include staying in character and drawing upon the skills and background of the class. After all, a PC is much more than a mere chess piece limited to a single type of move.
In this example, a fighter could draw upon the skills of his class in many ways. If the party were gathering information from underworld types, the scowling presence of a well-armed fighting man would not only keep the party safe but help intimidate their sources and induce them to cooperate. Likewise, he could guard their goods from thieves, take care of any local bullies in the tavern, and generally threaten force in subtle ways to advance the party’s cause. As long as the player of the fighter PC worked to stay in character this way, the DM would certainly be justified in awarding the fighter a full share of experience points, even if he never drew his sword.
I think this is quite thoughtful, but also suggests a pretty passive role for the fighter ("guard their goods from thieves", "threaten force in subtle ways").
I find it hard to envision a more active solution for the fighter that (i) preserves strong class siloes and (ii) does not involve metagame mechanics. Spellcasters have magic, which makes metagame abilities unnecessary. Thieves/rogues have the full range of skills - which, as per 4e (as an example), can be adjudicated in a way that puts them on a par with a lot of magic. But the fighter is not going to have either of these things (because then would be either a caster or a thief/rogue).
The 4e approach - of buiding core class features around combat, and then using skills and rituals, which are more flexibly accessed (if not completely classless), to handle non-combat - is an alternative, but establishes a fairly high minimum degree of focus on combat as an element of the game.