This is, unfortunately, the key sticking point between class minimalists and...let's say "medianists" like me who want a larger but still constrained set.
Specifically, they reject the idea that it "will just suck" as a subclass. They think it will be perfectly adequate. Indeed, they may even consider the subclass itself a superfluous thing, undertaken only because having nothing that mechanically captures the idea (rather than doing so purely thematically or..."procedurally" for lack of a better term) would annoy other players.
Of course, I'm completely with you on this. There are some concepts that will execute on both the mechanical and thematic concept purely by being a subclass. For example, I'd say the Warlock does a very good job for covering the "touched by an angel" thematic and mechanical concept via the Celestial subclass, which kinda turns the class on its head relative to other Patrons, but in practice it works quite well and actually makes an extremely versatile character who can do a little bit of everything (especially in 5.5e, where pacts aren't exclusive.)
But there are other concepts that are simply too big to be compressed into the thin, meager space that subclasses permit. This is especially notable with classes where the base class is already suffused with features, such that making a too-powerful subclass would seriously disrupt game balance. The Fighter is a great example here, in that its base chassis has...a lot. There's a lot there. I'm not personally happy with the lot that is there, but my preferences are irrelevant to the question of whether there is a lot there or not. As a result of those things, it's hard to (for example) make a truly "support-heavy" subclass, because the Fighter is inherently starting from a high baseline of personal survival and personal damage output. (It's still behind other classes IMO, but 5.5e is at least trying to address that...even if it doesn't always succeed.)
And then there are some ideas that you can do a lesser/incomplete version as a subclass of another class, but you won't get the full impact without a proper class. My everpresent go-to for this is the Eldritch Knight versus the Wizard, and then analogizing that to the Battle Master versus the Warlord (or "Captain" or "Banneret" or "Torchbearer" or "Tactician" or whatever one likes if the term "Warlord" offends). That is, we live in a world where the Eldritch Knight is derived from the Wizard; it is explicitly and intentionally a lesser equivalent, a pale shadow of what a true arcane-focused learned spellcaster can be, albeit with dramatically more personal defensive and offensive capability. The Battle Master is to the Warlord what the Eldritch Knight is to the Wizard: a pale shadow of what a true Martial Tactics expert would be, albeit with dramatically more personal defensive and offensive capability.
The point of the previous passage (not that you need to hear any of this!) is merely to show that even if you establish that there already is a subclass for a concept (and thus doubly so if you merely assert that you could make such a subclass), that isn't enough to justify axing the concept. Nature Clerics exist alongside Druids. Battle Clerics exist alongside Paladins. Bards exist alongside Arcane Tricksters. Wizards exist alongside Eldritch Knights. Etc. The mere existence of a subclass analogous to a possible class, or a possible subclass analogous to a real or possible class, is not enough to say that a thing definitely shouldn't exist.
Sometimes a 3pp comes up something novel that hadn't been thought of before in a previous RPG.
