Ok, how about I turn this around... What makes the "warlord" suitable for a separate class as opposed to being a subclass of a current class?
It's too broad a concept for a single sub-class. In addition, the candidate classes to hide it under are Bard, which has magic-use as a major class feature (not sub-class/'archetype' feature), and Fighter, which has high-DPR via multi-attack as it's major class feature (not reasonable to take away for an archetype. Neither spell casting nor multi-attacking for high DPR are suitable class features for the Warlord. Even if you decided that taking away major class features was an appropriate way to design a sub-class, the result would be tantamount to a new class, anyway.
We're talking about D&D classes for 5e... why wouldn't the fact that it's recognizable as a D&D class by fans of various editions not be a good reason to keep them separate classes?
That is the topic, but /point/ I was answering was your false claim that the warlord concept was not present in genre nor popular consciousness, but the Paladin & Ranger were (and, by implication, therefor the former should be removed from D&D and the latter two retained), with TV Tropes (of all the crazy things) presented as 'proof.'
I used the same source to demonstrate that the concept was very common in popular consciousness, /and/ that the Paladin and Ranger were primarily becoming so because of their long history in D&D.
Especially when one of the goals of this edition was to unify fans across editions? How many editions has "warlord" been an actual class in (not necessarily corebooks but editions)?
Exactly as many as it needs to justify being put in the Core Game:
one, in a Players Handbook 1. That it was 4e, the target of the malice and strife of the edition war, makes it even more critical to /include/ to avoid the appearance of 5e being an exclusionary 'h4ter's revenge' edition. "Unify fans across editions" does not mean "pointedly exclude fans of 4e." I know there were fans who felt excluded by 4e's sacred-cattle-mutilation, but two wrongs don't make a right.
I believe exactly one... even the Warlock has been in more editions than that.
Neither does "unify fans of across editions" mean "unify fans of most editions."
We're talking about classes not roles
No, specifically we were talking about archetypes in the regular-language sense. Classes often represent those.
... isn't "leader" a role?
Leader was the jargon label of a formal Role, yes. That jargon meaning evidently had no influence on the Leader trope article (which didn't mention or link D&D), nor, presumably, upon popular consciousness. Thus the Leader trope is clear evidence the fantasy (and broader fiction) archetype the Warlord modeled (as best D&D mechanics could handle modeling anything in genre, which has never been wonderfully well).
And isn't it covered by other classes already?
The Role? Not so formally, anymore, but to an extent sure. 5e isn't designed around neatly filling roles, and the classes that were formal 'Leaders' could no longer /strictly/ be called such. They may have some of the mechanical support, but it's more based on concept than role (and was circular, in the case of the Cleric, anyway, since the formal role had a foundation in the de-facto 'healer' role the cleric often fell, arguably, too tightly, into in the classic game). There are 38 sub-classes and only 4 roles, so there's a lot of duplication, and, so obviously it shouldn't need to be stated, no plaintive refrain of "but there's already a sub-class with that role" would be a meaningful argument against a class concept.
The concept, no. The Battlemaster gives a small nod to a sub-set - the 'lead from the front' style - but is far too deeply mechanically committed toughness and to high-DPR multi-attacking as it's primary party contribution, to have design-space available to also take up as substantial a support role (and as great a degree of flexibility) as would be needed to cover the warlord concept.
So again why should "Warlord" be a separate class?
I believe the above answers (and many other given up-thread) do cover that. But, really, the most basic reason is that 5e is meant to unify the fans of all prior editions, and the Warlord was a very well-developed, separate class, in the core-est of core books, the PH1, of one of those editions. Every other class that's appeared
unambiguously as such made it into the PH1. Even some classes, like the Paladin, Ranger, and Druid, that first appeared as sub-classes, are fully-separate classes in 5e.
My post was mostly musings. I think a more generic Rogue class for the first 2 levels (one that doesn't have Thieves Cant and all the embedded baggage that goes with thievery) and that can give out his SA as Insightful Strike 1/turn (rather than manipulating the action economy), give out Cunning Action, and do all the other Warlordy stuff (get medium armor proficiency and shield, skill dice, initiative bonus, let players expend HD during combat to "Rally" - Martial Inspirational Healing, have tactical movement for advances/withdraws/flanking et al) in Subclass might be nifty. But again, at that point, you are likely just as well creating the full-fledged class rather than trying to have the Subclass and a pair of feats do all the load-bearing.
As long as we're musing, the Thief has never been a terribly well-justified class and hasn't often meshed well with the rest of the game. In the light of some of the insights that Mr. Mearls shared in the run up to the playtest, it seems like a big part of the problem is that it started as the only class primarily - and mechanically - focused on the Exploration pillar. That changed some in 3.0 with SA becoming very potent, and 5e has stuck with that. But, without a very potent SA that could be stacked with multiple attacks, most of the Rogues goodies could be given to the fighter (or vice versa) to create a single class that would be strongly contributing in both the Combat and Exploration pillars, without being at all imbalanced overall.
I could see a less co-equal-party-dogma game that consolidated classes down to such a 'Hero' or 'Adventurer' class encompassing the fighter, rogue, and much of the more lead-from-the-front Warlord, along side a Magic-User class encompassing the truly capable, supernaturally empowered Heros, and a highly customizeable catch-all side-kick class covering support, niche, high-concept, comic-relief, and other secondary narrative roles. Such might even remotely work as a very extensive module.
My personal favorite aspect of the 4e Warlord (as GM) was the versatility of the Princess Build to accomplish various tropes. As I mentioned above, the tropes that follow from "Frodo/Bilbo as Warlord" (the plucky adventurer who is in so far over his head but "mans up" anyway couldn't be more inspiring...that can't help but rub off on companions) are vast in genre fiction and profoundly awesome in play (both as a PC and as a GM-side tool).
The whole 'reluctant hero' or 'in over your head' thing is incredibly hard to in the RPG context, because players really do have to be assertive and pro-active about involving themselves in the game and their characters in the story for the whole 'troupe style' or 'cooperative' thing to work. So was the Leader trope, because of the similar need for parties to be co-equal if not downright democratic in the gamist need for fairness.
Just as the Warlord slid Leader schtick in under mechanical support functions, I suppose the Lazy or Princess build and similar ideas or expansions of the class could do the same for seemingly non- or low-contributing character tropes, like the plucky side kick ('Caddy' as has been suggested), reluctant hero, in-over-his-head victim of circumstance, literal victim in need of rescue, critically important non-combatant being escorted, or whatever. Those mechanics, once developed for the Warlord (or some other name if the concept became even broader - I'd suggest "Icon," but 13A has made that mean 'campaign-defining-god-being'), could be ideal for NPCs, as well.
Yeah I think I've been looking for some factual reason or detailed explanation when really this is what it boils down to... And since I'm not necessarily looking for a Warlord type class for 5e, maybe it's less important to me... though looking over the Marshal for 3e is making me curious to see what a 5e martial-leader type character would look like... there are just so many other archetypes like psionic using characters (not to mention expansions to the current classes) that I want... that a new warlord class is not at the top of my want list... but I guess for others it's number 1.
Good news: Psionics (also cut from the PH, in spite of being in the 1e AD&D PH 1 - I know, appendix, not technically a class, blah, blah) is totally in the pipe-line, we've seen a development version and gotten to give survey-based feedback. It's clearly going to hit the Advanced Game before the Warlord.
Sounds like you have a "1st world problem" with the Warlord: you've been getting everything you want from 5e, before everyone else, and may realize, but not have internalized, that that's not true of everyone.
So, your issues have nothing to do with the actual class but rather what some player "might" do with it?
Have you actually seen anyone playing a 4e warlord do this? Has this actually come up in your game?
Lanefan never played 4e, so that's not really a fair question.
While I've tried to re-assure him it shouldn't be an issue because it was fine in 4e, I also have to say that it's an understandable misgiving, that, while 4e managed to introduce a 'Leader' Role and a PC class whose powers' fluff were often about actually providing in-combat direction, 5e, with it's somewhat more mingled fluff-and-rules-text might run into issues. While I'd share that misgiving if we just considered the 'RaW,' 5e's approach marginalizes rules problems by relying first & foremost on the DM ('rulings not rules'), so I think it's an issue any competent 5e DM could deal with neatly and easily, quite probably without even noticing it, since keeping one player from dominating the game is just part of the 'cat herding' we DMs do every session.