I deal with that by ensuring that there's no 5 minute work day. If you have 5-8 encounters between long rests, the spellcasters can't steal the show every time because they simply run out of options.
Or maybe it's just that I view it as a team effort and sometimes some PCs are going to contribute more. In any case, I enjoy playing fighters in 5E even if they're not as flashy.
High level D&D spellcasters have an abundance of ways to reliably dictate Rest Cycles (some of them Rituals).
If your answer to this problem is to artificially (meaning reducing the competitive integrity of the game to nothing) introduce balance by deploying GM Force (playing an adversarial, Calvinball game of preconceived arms race gambits because you’ve got unparalleled metagame access...proactively deploying blocks or, worse, impromptu blocks to ensure the spell gambit doesn’t work)...well, that is a non-starter for Challenge-based play.
But let us say for a minute you don’t do that and somehow there are organic situations that consistently, day in and day out, foil a spellcasters ability to deploy Rest Cycle dictating gambits.
8 (let’s go all the way to the max) resource-attrition conflicts per day (which, by the way, is an absurdly high number)...be it combat, parlay, dealing with travel obstacles, divination/investigation, what-have-you.
An endgame Diviner is going to have:
* 2-3 Divination plays (that will probably dictate an encounter with ensuring a failed save on a brutal spell or saving an ally/themselves from a bad spell effect.
* Potent Cantrips (particularly Minor Illusion)
* 20+ spells including Renewables
* Rituals
* Singularly powerful spells that outright win a conflict/obviate an obstacle by themselves
* Magic items
* Skills in the Int/Wis domain
Let us say, worse case scenario, they need somewhere around 3.5 productive (or decisive, as some spellcasters plays will be) plays/moves per conflict; 28ish?
Even at the far end of the spectrum, that is trivially achievable by an endgame Diviner. The idea that they have to meticulously ration their plays just doesn’t bear out in play.
Yes, early on, there will be some rationing. But the fact is that (a) the 6-8 (ridiculous) encounter model doesn’t proliferate, (b) Wizard resources both proliferate and increase in potency which outscals that model, (c) that potency means LESS moves required per day because “moves per conflict” will decrease as play progresses (with the wizard, and some other party members, attaining decisive “one-move-encounter-enders” in their repertoire.
So even if a GM Calvinballs Tiny Hut and the like over and over and over...the encounter model (even at its apex) is still not enough to introduce this theorized “wizards must carefully ration their resources!” paradigm.