I really like what Mike Mearls is doing with his video series...
This is the first I'd heard of it. It does seem an interesting peek into the design process, at least, a one-man version of it. I'm sure it's a little different in real development when he's delegating more of the grunt work, not just brainstorming by himself, etc...
For those that are throwing around adjectives like they mean anything
Adjectives are a perfectly cromulent part of speech.... ;P
, can you try to provide a proper thematic description for the class, it's core purpose, at least a framework of the mechanics, and in what way the proposed subclasses are unique enough to be worth building?
"Unique enough to be worth building" is a very, very low bar, if you consider that every sub-class to see print so far in 5e must have cleared it. I mean, how unique is an EK from a Bladesinger? A PDK from an OotC? A battlerager from a berserker? A War Cleric from a Paladin? Heck, whole classes are rather lacking in uniqueness - Sorcerer vs Wizard vs Warlock.
5e design, perhaps only as a side effect of trying to integrate so much that had come before or in trying to keep certain things optional, tends to give a player multiple paths to very similar concepts or even the same concept. A classic icon of early D&D was the Elf, able to fight and cast spells. An elven Eldritch Knight, in 5e, covers that concept. So does a SCAG Bladesinger. So does an MC'd Fighter/Wizard. A wizard with the Soldier Background and Martial Adept feat or a Fighter with the Sage background and Magic Initiate feat might not even be that far off.
And, while 5e, to the extent it's balanced at all (and after watching this podcast, I have even less confidence about that), is at worst spot-light balanced, it does not go so far as niche-protection. You don't "need" a Cleric to heal - Bards, Paladins, and Druids can heal quite well, too. You don't "need" a Thief to open locks - anyone can take the criminal background and pick up proficiency with Thieves' Tools.
There is tremendous overlap, both functional and conceptual, among the classes, sub-classes, backgrounds & feats.
To understand what's missing, consider if an existing class were missing, but it's overlapping bits were still there: Imagine there's no Wizard. It's easy, if you try.

But, there's still the EK & AT, there's still the Warlock and Sorcerer? What's missing? Couldn't maybe a Tome-pact Warlock or a new "Pure-Blood" (as opposed to muggle!) Sorcerer do the trick? What's so unique about a wizard? Books? Really, that's it? Anyone can surround themselves with some books...
Sounds pretty grim, but there's actually a lot that vanishes with the Wizard. The Wizard is the only arcane class that prepares its spells, the rest have a limited known-spells. There are 33 spells that'd vanish from the game with the Wizard - they're not on any other list. The wizard also represents something of an extreme point in the design, being traditionally the lowest-hp, least-skilled-with-weapons, most-magically-inclined of all the classes. And, it uses INT as it's caster stat, heck, it's virtually the only class to make significant use of INT.
As to it's practical contributions, yeah, it's a lot like a Sorcerer or Warlock (or Cleric, Bard or Druid), it casts spells a lot of them, of quite a variety, but it leans towards offensive & utility spells, and isn't 'burdened' by the presence of healing and other 'support' spells putting pressure on it's slots. So it's a Sorcerer/Warlock-type class in contribution, but subtly different in concept & mechanics.
You can see how the wizard, had it been in the game only a few years, might seem superfluous given all that, but, even then, it wouldn't actually be, given even a small toe-hold, it'd be unique enough.
To plug the Warlord into that analogy, there are bits of it sitting around. D&D without the wizard still has the EK and AT, without the Warlord, it still has the Battlemaster, which uses vaguely similar mechanics, and the Mastermind. Sub-classes that hint at a bit of the whole. The Cleric, Druid & Bard make similar contributions, but in very different ways both mechanically and conceptually. The Bard is closer, conceptually (the Bard even has a minor feature called 'Inspiration'), but where the Wizard in our analogy hangs it's hat on casting some different spells, and doing so using INT instead of CHA, the Warlord is unique from the Bard in not being a caster, /at all/, a huge conceptual difference.
That's it in a nutshell. A non-caster (non-magical, even, at least, at it's base, sub-classes in 5e can tack on that sort of thing), that makes primarily support-oriented contributions substantial enough to replace the traditional Cleric's contributions (or the Druid or Bard alternatives).
The game has elements that point to the ability of non-casters (even if entirely non-magical) to make support contributions: A few BM maneuvers, the feat, Martial Adept, that lets you tap one, the Inspiring Leader and Healer feats, the PDK sub-class doing non-magical healing, the Mastermind sub-class, and a few other odds and ends. But, it has no class that's primarily about that. Adding one - the Warlord, which did just that in 4e quite nicely - doesn't just open up character-concept options for players, it makes campaigns that de-emphasize or remove PC casting entirely suddenly much more practical.
I have literally never touched a 4E book.
Maybe you should? There's no small amount of stuff from 4e in 5e, and it might give you some insights. Same goes, obviously, for other prior editions, especially AD&D. 5e is very much shaped by the legacy and history of the game.
4e gave us 6 official 'builds,' plus an alternate feature, each of which would map to a 5e sub-class. I've already listed them, but, until you get the chance to touch not one, but three 4e books, I'll throw in a brief description of the concept, and how it might be done in 5e.
Tactical - a canny warlord, who excels at devising & coordinating cunning plans. This is the one that Mearls was talking about in the podcast as if it were the whole class, so, in 5e would use INT and 'Gambits' that map, vaguely, to spells in the way Mearls went into, only, to do it 'right,' it'd map more to the casting of Druid or Wizard than an EK. It would emphasize 'Tactical Gambits' in the same sense that an Evoker emphasizes blasty spells, an emphasis, being better at it, not in the sense of being unable to use everything else.
Inspiring - the original opposite number to the tactical warlord, the inspiring warlord did exactly what it says on the tin, bolstered his allies (hps, both healing &temps and handed out buffs), mainly keyed off CHA. It tended to be less about maneuvering & commanding and more about leading & aiding. In 5e, it would use the same Gambits & Maneuvers as other warlords, but better at the ones that hand out bonuses and hps, probably by the simple expedient of tacking his CHA mod onto them.
Resourceful - where the tactical warlord plans & orchestrates, the resourceful warlord reacts to opportunity - he has contingencies. The resourceful sub-class in 5e might be like the still unrealized generalist wizard, it'd be reasonably good at all the various sorts of Gambits, and it'd benefit a little, from both INT & CHA. Or, it could emphasize gambits that react to the enemy and leverage the environment.
Bravura - Mearls also hinted a this one, just a bit, in talking about the benighted pre-sub-class Warlord-heading fighter. It's the lead from the front type. It leads by example, is all about showing and inspiring bravery - and intimidating the enemy, something 5e could afford to emphasize more than 4e did. In 5e terms, this'd be the faux-fighter-multi-class sub-class, analogous to a war cleric, valor bard, or bladesinger. It'd get an actual extra attack of it's own, get slightly better weapons or armor, and do better with Gambits that involved getting in there and mixing it up to make things happen, rather than those that provoke or trick the enemy or direct or coordinate allies.
These next three were the last gasp of the warlord before essential (keep in mind, everything Warlord in 4e came out in a 2-year period, as soon as Mearls took the helm, the Warlord got nothing - except getting slapped with the 'Marshal' sub-class label) got less support than the others, so they're not as fleshed out...
Skirmishing - Emphasized mobility for both itself and it's allies. Skirmishing is RL military tactic, of course, in 4e, there were plenty of exploits that involved movement. In 5e, it could get a more evocative treatment, maybe emphasize DEX and light armor, and gambits best used by similarly mobile allies, and, of course, be particularly good at those gambits that involved maneuvering allies, quick in-and-out attacks and the like.
Insightful - 'Watchful' might've been a good name for this one. Specialized understanding the enemy and staying alert for their plans and actions. Mechanically it didn't much deliver, it mostly just subbed WIS for INT or CHA when handing out bonuses. In 5e it could get a more interesting "know your enemy" sort of treatment, shading over into what in 4e would have been off-limits 'controller' functions, and imposing conditions and actions on the enemy, metaphorically 'getting in their heads' and predicting & manipulating them. Could be very good at a few such gambits that are otherwise marginal in the hands of most other warlords.
and Archery - though the distinction isn't important, this was not a build, but an alternate feature ::shrug:: - 4e was not super flexible about the choice of range vs melee weapons, STR vs DEX, so the Warlord was mostly STR/Melee. This version was able to do ranged. It was the sole official exception to the 'lead from the front' idiom, it would shoot enemies and set them up for allies. In 5e, it'd be a lot less convoluted to emphasize ranged weapons, and an Archer-Warlord could simply do so, and excel at Gambits involving archery, his own or coordinating with his allies.
Then, there's the 'Lazy' build, it wasn't spelled out, but players strung together some tactical 'exploits' (maneuvers or gambits), that didn't involve the warlord attacking (or often, even acting, at all), into a build that aided allies primarily by funneling actions to them. The concepts this opens up are surprising - Garthanos called it a 'Princess Build,' because it could be used to let you play a seemingly-helpless side-kick or damsel in distress sort of character, yet still fully-contribute to the success of your party. Instead of imperiously commanding your allies to attack, you scream for help.

In 5e, this sub-class could de-emphasize weapons & armor, emphasize CHA and Gambits that involve heroics on the part of their allies.
Then there's all those Warlord-focused Paragon Paths, any of which might also inspire a sub-class. (Or a PrC. Have I mentioned, lately, that 5e could really benefit from 3.5-style PrCs? I think I have.)
There's 38 of them, in all. One, the Purple Dragon Knight, already a fighter sub-class (really, really, should have been a PrC, just say'n... OK, I'll shut up about it for a few minutes.).
For another instance, there's an 'Arcane Battlemaster' (I think Mage-Captain, might've been a good name), that actually did get a few (3) spells. That could be the EK-like faux-multi-class-Wizard sub-class. But the cool potential, with this sub-class is applying all those mostly-martial-focused gambits & tactics to casters. Free attack? How 'bout a free cantrip? How about I help you maintain your concentration...? Need a forward observer for that fireball?\
yeah. ;>
Some of them are race-specific (Spiral Tactician, Earthfast Brigadier), like the Battlerager in SCAG is, so would seem to be OK for a 5e sub-class.
Some are religious (Platinum Warlord, Dujun of Erathis, Battlord of Kord), they could be bundled into a pally or cleric faux-multiclass.
Combat Veteran was one I rather liked, the Path that gave us the grizzled drill sergeant vision of the Warlord.
etc... too much to go into, really...
But 5e classes aren't as focused as 4e classes were. The Warlord could stray into 'controller' and striker functions, as well, could literally lead bands of NPCs (perhaps under the sub-class name 'Marshal' as in "marshalling the volunteers"), and, of course, could go ahead and like the Fighter & Rogue, have a spell-casting sub-class or two.
The Marshal, BTW, was a Battlesystem class - not in the league of most 3.x PC classes, but better than an NPC class - that vaguely presaged the Warlord. It really was in a tactical miniatures game, and it's thing was 'auras' that gave fairly bland bonuses to allies in them. Yippee.
So, finally:
I like the idea of re-imagining the Marshal as a sub-class that focuses on larger numbers of (necessarily NPC) allies. Like the old-school Fighter and his (generally useless) band of men-at-arms he attracted for building that enormously expensive keep. But, y'know, not so generally useless.
Could specialize in gambits that involve the whole party, or the tired 'organize the villagers to defend themselves' trope. Would have been untenable in 4e because launching volleys of missle fire or setting up shield walls or whatever other modest-scale military maneuvers would have stomped into 'controller' territory. In 5e, it'd be fine to have some gambits like that, and a sub-class that was particularly good at them.