Thing is, once you've gone past the most basic things - OK, it's to be a medieval-fantasy-based table-top RPG preferably with party-based play to emulate the adventurers in LotR - the answer to "What is this game about?" is sooner or later gong to be "Within the parameters listed, everything."
The thing I like about D&D is it hits that "Everything" answer sooner than most bespoke games (put another way, it has fewer confining parameters), and then tries to cover that ground. Flexibility rather than pinpoint accuracy, sacrificing perfection in a narrow area for good enough in a wide area.
See, this is what is so odd to me about all the people who say this, I don't get it at all. Not just me, the whole history of the RPG industry, and the actions of TSR's designers, etc. have all pointed out the same thing clearly. D&D is a one-trick pony. Yes, TSR did once do a brief experiment with implementing another game on the 'chassis' of D&D, Metamorphosis Alpha, and its 2nd Edition, Gamma World 1. They're good games, but more DESPITE being based on D&D rather than because of it. And in terms of 'based on' GW is VERY loosely based on D&D. For example you get CON d6 hit points at level 1 and they never change after that. Leveling barely matters at all, and I don't ever recall that we even bothered to tally XP, it does practically nothing for you. The whole game progression is based on gear and/or lucking out and getting good mutations when you inevitably get fried with radiation. There are no classes, combat is only roughly similar to D&D, etc.
TSR never tried this again. Over the next 25 years they released many many games, exactly ZERO of which are based, even loosely, on D&D. I mean, why not loosely base Boot Hill on D&D? Its stats could have been written as 3d6 ranges. Heck, 'SPEED' could have been called 'DEXTERITY', and they could have included an INT and CON, though bravery and accuracy are kind of their own things, they still could be d6 based. CHA could even have been included if they wanted. But no, its a d100 based system, totally different. Also its damage and wound system are pretty important to the genre and combat wouldn't have been similar either, with its emphasis on gunfire, punch/grapple (something D&D has famously failed to do for 50 years now), and close in knife fighting (another weakness of D&D).
Same with every other game they came out with, Gangbusters, Top Secret, Star Frontiers, etc. No attempt was ever made to even leverage whatever overlap in mechanics COULD have been tried. Nor was this some failing of imagination, as BRP, and then GURPS, just to mention the really high profile ones, demonstrated that sharing mechanics was a workable concept.
I put it to you that the TSR developers understood, FROM DAY ONE, that system matters. That they understood extremely well that mechanics and story are intimately and subtly tied to each other and cannot be divorced, and that D&D is not a very flexible paradigm for most games. I mean, TSR's designers were ALSO constrained by the state of the art of games in their time, and had NO clue, AFAICT, that something like narratively focused games with story and scene framing mechanics were even POSSIBLE. At least they, and their contemporaries in the '80s, never concluded that a game could be built around those sorts of mechanics, ala today's 'indie games'.
I submit that the paradigm of levels, XP, increasing hit points, hit points and AC as a combat model, no or unstructured check systems, classes, and GM-centered referee-style story telling, are only applicable in that combination to a very narrow niche within the wider fantasy genre. WotC has shown 2 things in the last 20 years. One is that it is pretty hard to get that formula to work, even by hacking a bunch of pieces off it, for anything else (d20, particularly modern) cannot really be called much of a success. A few games that are very genre-adjacent to D&D itself were moderately successful. Some of the d20 modern games got some play, for a brief time, but in almost every case where such a game was successful it was re-implemented without using d20. Secondly WotC proved that you CAN create a non-D&D game based on D&D, the d20-based Star Wars, SWSE, game is a good game and seems to work well with a very D&D-like structure. I would point out that Star Wars itself, as a sub-genre of science fantasy, is almost custom designed to be D&D-like. Again, its a narrow niche, when you hit that niche, then D&D shines. Outside of that, not so much.
And obviously you CAN take a genre, horror, heist, secret agents, pirates, three musketeers, mystery, etc. etc. etc. and you can do a version of it within your D&D game for a time, as long as that version is pretty tightly constrained to what will work within D&D's overall paradigm. I mean, try to do a slasher style horror. Are you really going to kill a party member every 20 minutes for 2 hours? How is that mechanically going to work (I mean assuming they're not level 1 PCs)? The very 'armed gang of tough, competent problem solvers' ouvre simply isn't going to mesh with 'Freddy' running around ganking someone from the shadows, or 'off camera'. Plenty of aspects of such action that would benefit heavily from mechanics won't, because D&D lacks the appropriate ones. So, yes, we have Ravenloft, which is basically a Vampire Hunter drama, and that kind of works, though I think if you surveyed everyone who tried to run I6 you would find that it probably failed to an extent for most of them. The rest probably mostly ran it as a straight up adventure module and barely touched on the horror aspects.
D&D was a great start for the RPG hobby. Let it rest on its laurels. Its a good game, for what it does, and some of its elements CAN be pretty useful in other games of a particular sort, but to think that it represents a highly generalized model for RPGs that "just needs a bit of tweaking and some added mechanics" to make it a great match to lots of genres? No, sorry, the wisdom of 2 generations of RPG designers has now spoken on that.
