Are you trying to pass off opinion as fact here? Hope not...
I think "should" implies opinion or at least judgement in that context.
No you can't. At least not to anyone but yourself. You can't possibly dictate "what's missing" for others.
Now you may be inadvertently dismissing fact as opinion, it seems. (And Hussar, bless him, clearly has some deeply-felt opinions that are obviously shading his presentation of the facts.)
But, fact is, the BM has ~16 maneuvers, one of which bears the name of a Warlord at-will, but is short-rest recharge, a few of which map roughly to Warlord encounters, and the rest of which map better to 4e fighter tricks. The Warlord had it's Command Presence range of features and hundreds of 'maneuvers' (called 'exploits' in 4e), that the BM doesn't even begin to cover. That's fact. Hard numbers, not Hussar's opinion.
Whether and how you might think those numbers should be massaged and compared 'fairly' is a matter of opinion, but that the quantitative gulf between them is about two orders of magnitude is objective fact.
That seems to be the big disconnect in all of these threads. It is objective fact that everyone has a different idea as to what a warlord should look like. We see it in the posts, and we see it with all of the variations of the many warlord classes and subclasses people have created.
So what? Does anyone really have an idea what the Ranger should be? We've had multiple stabs at it, and it still seems murky. Doesn't mean it should never have been given a chance.
Therefore, it is impossible for anyone to say they can say exactly what is missing from what is published and present that as objective fact.
Actually, it's straightforward enough to see what's objectively missing from the current published options, since they are objectively there, in black and white in the PH & SCAG, covering 3 whole relatively feature-poor 'archetype' subclasses and a couple of feats, and the Warlord is objectively in the cannon in black & white, covering 6 official builds, the odd variation, and the CharOp lazy build, among other things, spread out over the PH1, MP, MP2 and the occasional dragon article, around a dozen class features, hundreds of powers & feats, and no small number of related Paragon Paths and even a few appropriate epic destinies & themes.
The sheer, objective, volume of what's 'missing' from the 5e not-Warlord is absolutely overwhelming.
BTW, to be fair, there's also the matter of what any one Warlord might be able to do. In 4e, the Warlord had hundreds of powers, for instance, so there were literally tens of thousands of possible, distinct, individual warlords, without even considering MCing, Feats, Paragon Paths, Themes, Backgrounds or Epic Destinies. In 5e, before considering the corresponding options of MCing, feats, & Backgrounds, there are exactly 3 (Any BM trying to play at being a Warlord will eventually choose all the Warlord-applicable maneuvers, the PDK & Mm have their warlord-ish features locked in, no choice involved).
But IMHO, it's not actually quite as bad as it looks on the surface: the same can be said for any 5e class, to a lesser extent. Even though there are hundreds of spells in 5e, each 5e full caster has only a relatively small sub-set of them that are unique to itself, and can, over 20 levels, learn most of them, meaning that they are defined mainly, as individuals, by the unique class spells they /don't/ know. Between sub-classes and unique spell choices, that's still scores if not hundreds of possible individual casters of each class (with the exception of the Sorcerer, with no unique spells and only 2 sub-classes + meta-magic choices for differentiation, of course)
Yet, at the same time, a given 5e caster can know over a dozen spells, and cast them spontaneously from, eventually, some 20 or so low-level slots, and a handful of high-level ones.
In contrast, a 4e character only ever gets 4 encounters, 4 dailies and some much-less-significant utilities, and can only use each exactly once. So, while 4e characters are theoretically differentiated from each other individually, 5e characters are, individually, much broader in capability and vastly more flexible in how they use that capability. If sleep is the best spell for the situation all day long, a high level 5e wizard can cast it a couple dozen times if he really stretches to do so, the 4e wizard, once, maybe two or three if he leverages very specific feat and magic item choices, and maybe slips a pre-errata trick past his DM.
In short, 4e traded a great deal of flexibility and effectiveness for tighter balance, greater differentiation, and role-support. Part of the huge gulf between the 4e Warlord and it's 5e
nth cousins is that difference in approach. To close that gap, the 5e Warlord wouldn't have to have hundreds of unique-to-the-Warlord maneuvers, it could, instead have only a few dozen, but have the ability master a fair majority of them, and use them with much great flexibility. A warlord proponent could still whine about the smaller number of possible warlords, but he couldn't doubt their effectiveness was improved at the individual level. And, it would adapt the Warlord to the 5e design paradigm, as evinced by actual 5e designs of other formerly-leader-in-4e classes.
Just out of curiosity, what is missing from a non-magical ninja in the assassin class? Surely it's just a matter of equipment, no?
IDK, could anyone in a typical fantasy setting make a smoke bomb or use a metsubishi effectively? The Essentials Executioner might have some ideas that could be incorporated.
Just some knee jerk reactions. It seems to me, from a very quick reading that a Shadow Monk fits the magic ninja pretty well and an assassin covers the non-magical one.
Much more coverage than the warlord, both in absolute and relative terms.
In absolute terms, the whole class and the classes' main thrust and concepts, not just bits of the sub sub-class, speak to the ninja in both cases: the rogue is sneaky, high-DPR, certainly fits the ninja, and the Assassin doubles-down on that, and, well, is an assassin; the Shadow Monk uses mystical powers to disappear into the shadows and power martial arts abilities (also all-ninja). So you have two base classes /plus/ two sub-classes worth of features virtually all of which are 4-square ninja goodies. In contrast, the main thrust of the fighter base class (tanky DPR, personal mastery of weaponry) and rogue base class (sneaky DPR and skill mastery) are at best side-lines of the warlord, and even the BM's maneuvers only have a few warlord-y options, so not even all the sub-class features apply - even expanding to SCAG, where most of the PDK and virtually all the Mm's sub-class tricks are at least a bit warlord-ish, they're still /just the sub-class/, and pulling against the main thrust of the base class rather than working with it.
(Arguably the monk suffers from the same phenomenon, just a bit: it may well be too all-in to the unarmed combat thing for a ninja, that's more nearly comparable to the problem of putting a warlord in a fighter archetype, though it's still much lesser in degree, just similar in type.)
In relative terms, that support need only cover the few extant concepts of 'ninja' - the mystic ninja, the highly-trained assassin ninja, the more obscure mountain-clansman-scout historical ninja (and, perhaps a Kunoichi, if they wanted to go there) - that are very narrow, specific, and culture-bound. The Warlord, OTOH, was used for a very broad range of concepts, not tied to a particular culture.
And that's the crux right there. I can pretty easily say what's missing from a Battlemaster if you want to make a warlord. Or a PDK. There's a rather large swath of elements missing to be honest and they've been enumerated numerous times.
It's actually pretty hard to spell out - it's easy to just throw out BM: 3 maneuvers, Warlord: 300, but to spell out how those might be translated into 5e (and expanded to mirror the greater breadth of 5e ex-leader-classes), is a daunting design undertaking, in itself.