5E is narrative first in that a player declares an action, the DM decides the narrative difficulty, selects the DC based on that narrative difficulty and then adjudicates a check. The DC comes from the DMs narrative judgement entirely.
Can you say what you mean by "narrative"?
In most RPGs, the basic process of play is that a player declares an action which in some fashion engages the shared fiction, and then a check (or comparable resolution process) is established and undertaken. Common examples are things like "I poke the floor in front of me with my 10' pole" or "I raise my hand in a signal of peace to the goblins" or "I cast a fireball spell".
In many RPGs, there are some actions whose declation in the fiction more-or-less
equates to performing a mechanical move in the game. In Gygax's AD&D, for instance, declaring "I listen at the door" enlivens a particular mechanical process (which contrasts with "I poke the floor in front of me with my 10' pole"). In 5e, declaring "I attack the goblin with my sword" or "I shoot the goblin with Magic Missiles" enlivens a particular mechanical process (the combat rules, and the casting rules, respectively). In most RPGs, there are also actions whose declaration enlivens no particular process; or envlivens a very generic process like "The referee determines the outcome" or "The referee will determine a throw that must be made." Moldvay Basic and Classic Traveller in particular have a lot of this.
4e's generic process is a bit less open-ended than Moldvay Basic and Classic Traveller: if the referee determines that a throw must be made, by default it will be made on a d20, will involve adding a stat or perhaps skill bonus, and will be against a DC selected from a DC-by-level chart. What makes it
fiction first is that the mechanics establish an abstract structure (at this PC level the appropriate DC is such-and-such) but don't peg any particular fiction to that structure. The fiction is given elsewhere, prior to the mechanics. This is why you can do stuff with 4e that can't easily be done with Moldvay Basic, like compression the default paragon fiction into the mechanics of upper heroic (which the Neverwinter supplement did) or extending the default paragon ficition into the mechanics of epic (which is what the Dark Sun supplements did).
To me, 5e does not seem to present an abstract mathematical structure onto which prior fiction is then appeneded. It doesn't have the mechanisms that I associate with such a structure, which in 4e are the DC-by-level chart, the corresponding creature build charts, the notion of "minionisation", etc. And it seems to have the mechanisms that I associate with a pegging of particular fiction to particular mechanics, like creatures whose mechanical specification is constant across all levels, and DCs which seem to be presented as "objective" rather than "subjective" - which in this thread has been reinforced by the suggestion that the way you gate something against a 1st level PC is by setting the DC at 27: that's mechanics before fiction, not vice versa.
The game is agnostic as to what is narratively appropriate in a campaign.
<snip>
The narrative is determined by the DM and table, the rules follow.
By "narrative" here you seem to mean "genre" and other associated aspects of fiction (tone, tropes, etc). But that doesn't go to the question I am asking about adjudication. Again, and to reiterate, if the way you gate things against 1st level PCs is by setting DCs that are mechanically impossible foe them, that is mechanics first, not fiction first. The mechanics are determining the feasibility of that action declaration, not the fiction.
If the DM and table don't think a particular challenge would need to be rolled for by Konan the Kongqurer, then the book advises not asking for a roll. DM judgement.
That's not the example at issue, though. No one has suggested that the way in which 4e empowers upper level martial PCs is by allowing the GM to say "yes" without requiring a roll. The dicussion has been about the methodology for setting rolls, and in particular whether the DC set for Konan's player is the same as the DC set for the player of a 1st level fighter.
A DC 25 task is very hard for low-level characters to accomplish, but it becomes more reasonable after 10th level or so. A DC 30 check is nearly impossible for most low-level characters. A 20th-level character with proficiency and a relevant ability score of 20 still needs a 19 or 20 on the die roll to succeed at a task of this difficulty.
To me, this text appears to imply that DCs are "objecive" rather than "subjective" - mechanics first rather than fiction first in the sense I have set out in this post.
what can a 15th level fighter do that a 1st level can't... the answer being whatever the DM, informed by the fiction, decides.
In 5e it's clear that if there's no chance for an action to succeed then no roll should take place.
In 4e if there's no chance for an action to succeed then no roll should take place. That's not a very distinctive princple.
But as far as
the GM decides, I'm asking what principles are expected to guide a GM in this respect. The main answer this thread has supplied is
martial PCs can't do supernatural things. This has been reinforced by references to bounded accuracy, which imply that DCs are "objective" so that if the GM sets (say) a DC 20 for the 15th level PC then that is what the DC would be for a 1st level PC also.
Am I wrong? Are there actual play examples to be given where the action was judged impossible for a 1st level PC but DC 15 (and so definitely feasible) for a 15th level PC?
just use 4e rules for both this and flanking. 
