Correct.
I am explicitly saying that I've come to the conclusion that the Wizard trope is a sacred cow of mainline D&D design.
That is an impediment to making the game better, yes.
I suppose your position is that it's an insurmountable one?
I can't really strongly disagree. The chance that WotC will try to make the game less terrible, when being terrible appears to be delivering revenue growth are nil.
But, how a game might be made better still strikes me as an engaging subject, no matter how academic
. The "treadmill" complaint is about the nature of the gameplay loop. Players deploying similar abilities to resolve problems in similar ways, thus that all that changes in the game is the magnitude of numbers and the kinds of description employed.
I mean, the 5e gameplay loop is the same at all levels (without much variability in the numbers, even), you don't hear "treadmill" too often (though Micah /did/ just say that about 5e & 3e, too, so I guess, I've heard it at least once). The classic D&D gameplay is resource management, you get more hp, spells, and items to manage, but that's as far as it goes.
The alternative is a game in which the actions available to players are transformative over time, thus that problems are not resolved the same way as they increase in level, and the kinds of narrative the game produces are different at different levels.
So, an example of a treadmill would be if the characters got some basic options in combat (attacks) and out (skills) that were numeric and rested on the same sorts of (d20) checks, and the bonuses and target numbers just edged up, perhaps you got to make the checks more often, but new options didn't open up.
So, for instance, the TSR era fighter or the 5e Champion fighter.
A not-treadmill would be if the characters gained new abilities as they advanced. Like all spellcasters in every edition, or 3e fighters gaining feats with new feats becoming available at higher level or with the acquisition of prerequisites, or 4e fighters gaining more and different exploits, which open up as they advance ... or 5e BM fighters if they had some better maneuvers that unlocked at higher level.
The treadmill assertion, as you put it, is about the means by which that state is achieved, not about the idea of evaluating encounters. Or at least, it shouldn't be
They certainly seem closely related. The treadmill assertion has always been that advancement is illusory, that the same moves and the same rolls will have the same results regadless of the level at which it's happening.
That speaks directly to how the DM calibrates challenges, and is not impossible in any version of D&D. A 5e DM, for instance, could, thanks to BA, very easily manipulate ACs, saves, and other DCs to make everything the PCs did have a 50/50 chance of success at every level. The better the tools the game gives the DM, the more easily the DM can set challenges to a desired level. In a game that leaves the DM to his own experience and creativity, the players will probably experience quite a range of challenges, even if the DM is trying to be consistent.
But, if you get more moves that do different things as you advance, that's less of an issue. 4e was only different from other editions that way, in not giving vastly more new moves to casters relative to non-casters.
The issue becomes more obvious in the more anemic non-combat systems, which are mostly just "roll with a 70%ish chance of success, with some variance," and that gameplay loop never actually changes, just the description of what is causing the player to roll.
... if we're honestly, kind of the design ethos that's been present in D&D adventure products for a long time, just not one well supported by the assorted core systems.
Non-combat has always been poorly supported in D&D relative to spellcasting and combat, usually boiling down to a single check to resolve a given task. Whether that was a Thief's % 'special' ability, or a WotC era d20 skill check, with modest exceptions like opposed checks, or the 3e complex (more than one success on the same skill required) skill checks, 4e skill challenges (multiple success on different skills from the whole party required), or 4e & 5e group checks (whole party rolls the same skill, half need to succeed).