It's popular in a variety of circles to entirely reject the "Combat is a Failstate" maxim, to deny it having any legitimacy. People do this for a variety of reasons that usually boil down to being unfamiliar and/or unhappy with the state of much Post-OSR and OSR design.
To understand the Maxim and all OSR maxims one needs to understand a few things:
A) Maxims are not axioms or universal truths.
They are pithy generalization that don't apply to all situations, but state or offer an example of a core principle. They tend towards making bold memorable statement. Finding a hole in a maxim or a counter example is like finding the same in a metaphor. The clouds do not literally roll like the ocean for example ... they are not fluid ... but you understand what I mean when I say it and would find someone saying "NUH-UH the clouds can't roll like the ocean because they are gas not liquid!" unhelpful. Though of course if the sky is still as a frozen lake ... the metaphor could just be wrong.
B) OSR Maxims in particular serve another purpose besides offering vague generalities about OSR play.
They are handshakes or code words that helped people recognize others interested in OSR style play and helped form group identity. Now that the OSR has fractured into many (often antagonistic) sub-groups (The Post OSR), and not all have adopted all the maxims (and no one ever did) so they become more useless and contentious.
C) The argument of "In 1974 - 1983 (or whatever early period) we didn't play that way" is invalid related to the OSR.
The OSR was not active in those times. The OSR was a style of play inspired by early game systems - and while a minority of its participants may have sought to emulate early play, most didn't really. Take one of the early formative OSR texts - Philotomy's Musings. It doesn't try to say "This is what Gygax meant in OD&D" it looks at OD&D as a "new" document and tries to make sense of it intrinsically without Gygax's later works or interviews and such. It ends up coming up with some OSR principles such as the "mythic underworld" (The old school dungeon is a hostile alien place that follows its own rules ... because as designed in OD&D the size and the minimal level of detail mean that it lacks coherence). These may have some resemblance to the way some tables played in the 1970's or 1980's but that's because it's applying the same source material. Remember that the West Coast style of 1970's play exists as well as the Lake Geneva style. One led to "Trad" games and 90% of modern RPGs... the other led to Tournament Modules and lawsuits over royalties. The OSR was not either of these styles and evolved it's own approach using old rulesets. The Post-OSR is fighting over ownership of them and often seeking to revise them - with some Post-OSR groups wanting to claim ownership of the OSR name.
...
On to "Combat is a Failstate"...
Personally I would say it's an "Comabt is and inevitable failstate" - Dungeon Crawling in the OSR style is like playing Jenga. You know the tower will fall at some point, and it's in the hands of each player to simply rush the process. The distinction is that in OSR style D&D (or quasi-D&D) combat doesn't automatically end the game, the maxim isn't "Combat means you lose". Instead, it's not optimal, it's a state where loss is possible and more likely ... it takes some measure of control away from the players and puts them in the hands of the dice and mechanics. It represents a higher risk and that level of risk varies - including with on how will the players set themselves up for the combat - i.e. the "Combat as War" maxim.
This works because OSR play and dungeon crawling most explicitly is not set up like modern D&D, it leans more on exploration then on tactical combat for the majority of its play. Yes you can run it as a skirmish game - and it seems like many of the earliest tables in the Midwest (both Twin Cities and Lake Geneva) did for a time. This is a perfectly valid playstyle. I play a different style of game (usually).
However, the OSR, or at least the portion of it that I interacted with from 2011 - 2020 ran games focused more of "fictional positioning" and "fruitful voids" -- that is treating the game as one of problem solving where foes were generally expected to be at least the party's equals, and a head to head fight would likely result in character deaths. Fights occurred (and still occur) in most game sessions, but are unpredictable and players rarely rush into them. This is the heart of the Maxim - which certainly overstates its point for maximum effect ... but it's a major point in this style of game and one of the ones most alien to 3.5E - 5E D&D, where combat has a lot of fun player tools and takes up large amounts of play time - it is often the locus of play.
That's my take, and I'm pretty sure it's Clayton's (who wrote the original post) as well. I currently run a dungeon crawl using OD&D rules with a party that regularly includes 7-8 players and 7-11 PCs and henchmen and the try to avoid combat when possible (or cheat it in their favor) because it has proven quite deadly. So this does work as a style of play - you just need to offer other and dangers obstacles as well as enemies that want to tussle.