I disagree. The mechanics don't have to be encounter-specific at all. Rather, the game is in desperate need of universal mechanics that can lend themselves to nuanced usage. I believe enmity, if revised, could be one of those things. If all mob attacks were skillshots, of sorts, and thereby interceptable (albeit it with whatever pierce/splash/cleave factor), that would be another. And that's barely scratching the surface. For instance:
"Stagger" as a nuance-capable party survival tool.
Since 1.x I'd been pushing for a revision of the XIII Stagger mechanic. Consider all damage as the product of Potency and Pierce. In XIII terms, this is the raw, initially Ravager-like effect of Stagger and the factor by which it can be modified towards the longer-lasting Commando-like effect that keeps it from draining as quickly. Jobs are balanced around having either more Potency or more Pierce by which to put out their intended damage levels. Those with the most Pierce, such as MCH, BRD, and DRG, are good at longer-term harassment of enemies through Stagger. Those with the least -- and therefore the most potency to compensate -- are good at immediate, short-term Stagger. But, this doesn't lead to an incapacitated effect. Instead, it simply directly saps from a combination of the enemy's Potency (based on the Potency of your attacks) being generated or readied at that time, draining over time and upon the drained skill's use, and Strength/Dexterity/Vitality (based on Pierce damage). In this system, Attack Power is integral to means of defense -- without it, you cannot significantly block or parry.
In this system, active mitigation would also be based on readied potency and upon one's main stats. Being bashed about reduces the strength with which you can block or parry right now. Being gutted reduces the main stats by which you can increase block/parry strength, block/parry chance, and your maximum HP and HP restoration over time and from external sources. Being overwhelmed will kill you. The advantage of tanks, then, is less that they can survive a single blow or flurry of blows that otherwise would not be able to, but they can continue to take hits without being trampled, and with the least negative impact on their own damage (proportionately, and perhaps even flatly). This allows non-tanks additional lenience by which to participate in tanking tasks, albeit only momentarily, and reason to do so. Inversely, the tank's survival depends largely upon the party focusing down and/or suppressing enemies who would do the most harm so that they tank can survive. Survival is now everyone's concern.
The key here as it applies to the present state of tanking -- party irrelevant -- is that not all skills Pierce equally. Enmity skills, therefore, would be replaced with skills highly capable of immediate Stagger, while everything else is invested with utility that allows you to tap into your resources, ready yourself, ramp higher, or whatever else possible in the time that you needn't bash an enemy back just before their blow so that you can survive by bare hundreds of HP. For Warrior, Skull Sunder has the most stagger immediately producible. Butcher's Block has the most over three hits, sustaining the Stagger but spiking for less. Storm's Eye peaks at the end of its animation and adds Armor penetration. Storm's Path sustains, heals, and has twice the gauge benefit of any other skill. So on and so forth.
Debuffs and DR.
Replace Diminishing Returns. Replace most pure Stuns/Silences. In their place, create spectrums of debuffs, with multiple tiers therein, inflicted with debuff "damage", so to speak. Stun degrades to Pacification/Silence which degrades to Slow (Action Speed). Bind degrades to Snare* which degrades to Heavy (Movement Speed). (Tugs, slowing heavily rather than completely, like trying to pull free of a trap, taking effect over displacement rather than breaking instantly upon movement again becoming possible.) Essentially, players universally and mobs by rank, type, or uniquely are given initial and end thresholds (think, static vs. kinetic friction coefficients) for a (de)buff of a given spectrum, be it Action (Haste <> Stun/Pacify/Maim), Movement (Alacrity <> Bind/Snare/Heavy, Vertical (Launch <> Smite), Incapacitation (Sleep/Apathy/Daze), or whatever, for when the spectrum will allow the next tier of effect. If a player fails to deal enough Action Speed-type debuff "damage" to meet the start threshold for Stun, it'll be degraded to whatever tier it can afford.
This debuff "damage" works on literally anything, but, naturally, at less effect the more mass or HP or resistance or whatnot an enemy has. Moreover, each time a debuff of a given debuff "Damage" amount is inflicted, resistance is added proportionate to that "Damage" over a second or so, with drain speed being inversely proportionate to the %effectiveness of the debuff that created it. A Stun has 100% effectiveness in preventing actions and Bind in preventing movement, so their resistances do not decrease while Bind is active, giving it a longer period during which some portion of Stun or Bind "Damage" is reduced. This means that debuffs, e.g. Stuns, are hard to stack continuously, because the more -say- Stun you hit for, each following stun is worth cumulatively less until the resistance has had time to wear off. This is your DR component, but done flexibly.
Let's say one player tries to stun a mob with some 1500 Stun "damage", but its threshold is 2000, then it instead inflicts Pacification and drains continuously based on the damage generation (e.g. building continuously towards the next special or two coming off cooldown but locked out due to pacification) being prevented. Let's say the mob builds 20% resistance, so 300 debuff damage mitigation. If another ally a second later hits the mob with another 1200 Stun damage after 200 has already drained and 280 is reduced by resistance (20 faded over the second), then we're looking at a total now of 2200, enough to inflict the full Stun. The Pacification is then consumed instead as a Stun, and the debuff Action Speed debuff "damage" is now consumed at the higher drain rate of a Stun, during which resistance does not fade.
Again, just as this works for players against mobs, this also works for mobs against players. Tanks will have higher thresholds, drain rates, and inflicted-resistance modifiers inherently atop just taking less debuff damage in the same way they take less HP damage.
They don't have to be, and so long as PLDs remain the best at that functionality due to some unique capacity or extent of their arsenal, they have nothing to fret over.
For instance, give them the best cumulative mitigation factor. Turn Cover into Protector, a trait by which allies behind you receive your Defense towards their mitigation against cleave attacks and cleaves dealt against/through you are doubly reduced in their damage to others after you by your mitigation.
Have cleaves deal x% damage dealt to the previous target to the next. Each target hit applies their Defense or other forms of mitigation and thereby reduces what's dealt to the ally behind them. This adds up cumulatively, e.g. over a Dullahan greatsword swing's worth of allies from tankiest to least when locked into an area from which one cannot job -- squishy healers at the far end so they can survive. A tank naturally takes less damage and thus passes on less, but a PLD in particular maybe affects cleave damage's cumulative reduction doubly when blocking, atop loosely reducing cleave damage taken by allies behind them. Now you have more to do as a tank, including as a Paladin, but Paladin remains special. If need be, replace the current Intervention with a stronger Aegis Boon spell-shield by which to block for another as if you had blocked it yourself, or with an ally-targeted charge that does action intervene -- intercepting the strike directly for your ally. Cover (now called Guarding Force?) not only allows you to intercept from a range, but allows you to absorb debuffs on the ally's behalf. Passage of Arms applies up to the whole of your mitigation to those behind you and lengthens the cone of effect, building off what all tanks have, which PLD naturally builds off even further, and now making some ridiculously strong tools out of those aspects.
Yes, but not because they need to be. Don't mistake an already criticized status quo with a fundamental truth.
I wholly disagree. I'm not recommending taking control away from the tank, nor even requiring considerable coordination within casual content. But, I feel far, far more immersed in and engaged by my role -- be it tanking, damage-dealing, or healing -- when my party is coordinating with it, when I know what they want from me and they know what I want from them, when the WHM knows to allow me the few moments of no-CD bashings and heals through it with only an Aero III to spare while I'm milking my Blood Price and The Blackest Night before I drop Grit completely in favor of a Blood Weapon DA-Quietus spam as the Holy Spam goes out and I've 7 seconds of complete mitigation before having to pop my own tool -- LD if sync-able with Bene -- everything else if not. It feels great when melee are already stretching to the edges of their hitboxes as you silence a caster and pull him out of where the incoming AoEs would be. And that's the kind of stuff all this would double down on -- the feeling of teamwork seen through a particular role, not because you have sole proprietorship over its functionality or responsibility, but because you have a toolkit that wants to be abused and by golly you're going to milk it for all its worth in the party- and encounter-context you're given.
I have to disagree again. They do indeed have the lead control over mob positioning, but if a tank honestly feels as if they personally control everything in the fight... they're probably not a terribly good tank. Granted, I may just be misreading your tone here, but as I see this far too often from... dare I say, "entitled" tanks, I have to point out a tank's control is ideally only to do what is best in the context of the fight and who's involved in it. Playing well, a tank doesn't exactly "get a say" outside of perhaps pull size. If Mythic+ dungeons or the like were a thing, where time actually matters beyond just not needlessly wasting others' time, even that wouldn't be the case. A great tank looks at the situation, and responds accordingly. There's no real personal "control" or expression there. There's just competency and responsibility, shown in a way that's a lot more visible than the other roles' in many a situation.
What makes a tank a tank is what mental processes are involved in playing it well. Memory. Awareness. (Of both mechanics and, in less structured content, one's teammates.) Planning. Procedure. (Since you can't wait and think about it once already under pressure.) And the better a tank gets at those things, the less they feel in control, yet the more they feel their able to adapt. There really shouldn't be any sort of 'party lead' grandstanding or similar feeling of role-play in being a tank. You just have a toolkit, same as anyone else, and you put it to best use.
Now, am I suggesting we make some of those mental processes less unique to being a tank? Yes. They're fun when done well. Letting others in on that fun outside of the occasional Savage gimmick lets more people feel that fun. And letting more people experience it doesn't take away from the unique perspective a tank's toolkit places on that fun. Control shouldn't be zero-sum with one's party. It should be without any limit-by-competition, and felt as an ability to respond to content. Coordination is fun. Competency is fun.
I've not suggested anything remotely sharing trajectory with that trend. Quite the opposite. I realize you may be waxing against some generalized faction of opinion, but it is not mine, and I would caution assuming such exact connections off a single sentence asking for "more tank-like things to do" (albeit not through gutting mitigation or enmity to force their 'importance' down our throats).
In the end, you have mechanical depth or you keep the shallow slew alone that we have now -- enmity slapping and circle-shirks and CD timing charts, no matter how you change the mitigation or the enmity generation. But depth of undermechanics has no sole recipient. Nor does it need to or ought to.
Your red herring has in turn gone and chased something shiny off into the abyss here...
My suggesting that DPS and healers might benefit from being encouraged to share in the aspects of awareness usually required only of a tank, to be more viably or truly given the ability to take part in positioning mobs, managing enmity, or keeping the party alive through reduced party damage intake, is not remotely like what you're suggesting above. We already have the things your indicating as a slippery slope here. They're irrelevant to improving the attractiveness of tanks or tank gameplay, but they're quite wholly already there. I am not asking for more gimmicks of interdependence. I'm asking to move away from gimmicks towards real manageable depth.