An excellent discussion of how a game can succeed despite some of its characteristics and not because of the whole set thereof. Indeed, one could argue the same thing happened to 3rd edition: the first 6ish levels were playtested and reasonably balanced, but after that it goes pretty quickly off the rails. Yet people loved (and in many cases still love) 3.x, its enormous mechanical variety, its nooks and crannies, sometimes even the ridiculous gonzo you could produce within its rules. They got hooked by the initial onboarding and were more willing to try to fix it than to move on to something better-designed, in many cases despite explicitly knowing and recognizing the many issues it has. That is quite literally why PF1e exists, and why PF2e wasn't an absolute knockout when it launched. People in this thread generally agree about how flawed and problematic 3.x/PF1e were, yet that edition family remains fairly popular (albeit much more niche than before.)
Even in this thread, we've had pro-5e folks admit that high-level 5e isn't great, that there are classes and subclasses that aren't well-made, that the DMG is below the standard it should have met as perhaps the most important book for D&D's long-term health, that they see certain default rules as seriously flawed and in actual need of replacement with official options or homebrew, etc. If we all had the ability to make choices in a perfectly logical context with no sunk-cost thinking and no emotional attachment, there's no guarantee at all that these issues wouldn't be enough to drive some folks (surely not all, but surely not none either) into seeking other games. And things like Level Up exist, and have rather significant popularity, because people recognize that 5e as it exists is incomplete, lacking support for some archetypes and making a poor showing for some of the archetypes it does claim to support.
And that hits on a key sticking point. How much change can you make and still call it “5e?” If you houserule half the core systems and use Level Up and rewrite several classes and all spell lists, as we know at least one person on this forum has done (well, maybe not the LU part, but DND_Reborn has not been remotely shy about discussing the dramatic and radical rewrites they've made), are you still playing “5e,” or are you playing your own OGL homebrew system which happens to resemble 5e in some ways?
I know this is a Ship of Theseus/sorites paradox situation, there are no clean answers. But the point stands that even among people who love 5e and have been there from the beginning, there is already some disagreement as to what 5e is or should be. To extrapolate from 5e's success to the idea that all of its systems are "near-perfect" or that it was truly, uniquely special in a way that is more important than the context in which it appeared...it just isn't justified.