That's what I mean. If I have to alter the stats of a monster to work in a normal way, then it "doesn't work".
Despite the fact that I make house rules, I consider every rule I make that isn't purely about aesthetic or thematic preference to reflect a design flaw of the system. I do not agree with the angle of D&D being a tool-kit and a suggestion box. RPGs should be fully functional experiences right out of the box. If I buy a car or a jacket or a sandwich, and I want to add a better stereo, sew in an extra pocket, or add horseradish, that's preference, not a design flaw of the product I bought. If on the other hand, I have to replace the wheels to make the car drive straight, sew on a zipper or buttons because the ones it came with break the first time you use them, or add condiments because the sandwich comes with none and lacks flavor, that is a design flaw in the product I bought. I consider D&D to have several of the latter type of elements.
I consider this a minor design flaw, because the learning curve of going from casual player to experienced player seems rather trivial to me, meaning that the experience seems geared to the "starting player". Now, maybe the majority of D&D players are casual players who never increase their skills, but that sounds doubtful to me, especially given that people seem to be reporting a lot of optimizers in AL. (I'm not making a binary distinction between casual players and optimizers. I think the majority of players fall into a middle category.)
It seems to me that a game ought to be designed to work as written for the average player, not the starting or expert player.
Re the bold, we have extremely different expectations for what "does work" means for an RPG.
Does chess "work"? By that i mean that if any two players of any two strengths sit down and play it by the rules they will get a competitive match? Chess is a simple board game with a stock setup of pieces and absolutely defined playspace etc - surely that can "work", right?
An RPG is not a "board game" type thing and its objectives are not to be like it. As such a *lot* more choices are given to the players and a lot more is expected of the GM. Its the nature of the beast. There is no prescribed formula for "this will be that challenging" just a loose CR set of guidelines which go into how they are just guidelines.
its an impossible task to an RPG to try and have it so that every monster as presented "works" (by whatever level of threat or endurance or interesting you want) for every group it might be able to face in every circumstances - so they do not actually try. Its all left to the Gm to determine - not dictated by the game in any way. They provide some guidance but not at all a defined setup that could even imaginatively be construed as "the way" for all the various and particulars.
If altering anything's "printed stats" to better meet the particulars of your game, your players, your preferences is to be seen as a failure, then no TTRPG ever made can be seen as a success by that standard.
Do you know of any TTRPG where that would not be true? Can you name three perhaps that would pass that test for say 50% of the purchasers?
If there was such, it would IMO likely be one of the more abstract indie ones where the basic play and design were very much radically different - like maybe say Screentime where the "game" is mostly absent and the play is more about shared and collaborative fiction and there are actually little to no "stats" presented. In other words - about the only way i can see it working, something passing that test, would be to not present any stats at all (other than perhaps scaled "PC+" and "PC-" scores in abstract. )
I often find it valuable when having discussions about "works" and "not works" on this scale to get some concrete examples of other similar systems where "works" is clear so we can get a handle on the goalposts being talked about.