Quote Originally Posted by ItMe View Post
I don't know man, the trinity LONG predates FF14 and has been used both successfully and poorly for almost, what, 3? 4 decades? I don't see how using it (not even in a particularly extreme way) is an inherently bad idea.

Also, correct me if I'm wrong, but it seems like you're talking about being disatisfied with the trinity, when your main issue is actually the overemphasis on job balance and a damage centric design philosophy?
The two come hand in hand surprisingly often, as, historically and in practice, the designing around a trinity tends to do more to trim modes of available responsibility or engagement than add depth to their interactions. That's neither necessary nor ideal, but that is typically the way things turn out, typically shallowing out the means of engaging with enemies to meatshield-and-enemy-number-drainer, ally-number-pumper-and-enemy-number-drainer, and purer-enemy-number-drainer, across respective Blue, Green, and Red roles.

That's not to say pure "generalist" play tends to be much better. Without any semblance of reliable tanking (which is different from a reliable tank), for instance, it is much harder to situate mechanics like positionals or directed baiting on skills/attacks that don't have precise conditions (like always goes for the furthest or closest player). Similarly, if no one can opt into increased eHP or the like, however temporarily but nonetheless at some cost, there is no point in designing ways to leverage that eHP into "margining" or "traded" exchanges in combat, which can otherwise be incredibly exciting (especially when they're not sustainable in the long-term). That, again, is an extreme, though. Some of the best (most involved, most complex) tanking experiences I've had have been in games without any dedicated tank role. Hell, it's almost felt like the more dedicated the role, the less tanking (in terms of actual gameplay, challenge, decision-making, etc.) is actually involved.

Consider, for instance, two ways of managing Enmity, between a trinity and non-trinity game, respectively.
In the trinity game, the answer is almost always simple: Enmity is simply damage x multiplier, and the multiplier is given (almost always exclusively) to tanks. Thereby, a tank basically performs the exact same gameplay loop as a DPS, but inherently hold threat (or "tanks").

Now consider the options a non-trinity has available to it. The easiest solution is to simply slap on a modifier, as above, which can be generated by stance or buff. Consider, again, though, the needs of a generalist game, especially for solo play, wherein snap mitigation tools are necessary, but enmity-increasing tools are not. As such, the tank stance may or may not provide increased eHP, let alone sustainability (the difference between mere increased max HP and mitigation). In either case, though, that eHP bonus will not likely make those snap mitigation tools unimportant. Therefore, tanking doesn't become the duty of just one person. While one player may be better at controlling the mob and therefore facilitate his team's efforts by holding attention more often than anyone else, everyone will still need to cycle through with their mitigation, whether by stacking split-attack damage, taunting (or, far harder, timing flexible burst damage to draw focus) and baiting or shielding, etc., etc. Largely by necessity, there's just far more involved. Rather than the tank existing so that everyone can mostly play mostly like mere DPS (the healer doesn't have to be particularly preemptive, as opposed to when healing a DPS; the tank doesn't actually have to do any tanking beyond a quick hit per mob, etc.), tanking exists for everyone to enjoyably engage with.
Now, while we might call a given game "trinity-based" or "generalist", in actual practice, it's more a sum of depth and breadth. If there's too little deliberate capacity, it's hard to situate related mechanics and thus depth falls off. But if tanking truly becomes the domain of only one role, thus trimming available gameplay from everyone else, then breadth has been sacrificed.

In WoW's M+, for instance, tanking will tend to feel more like a group effort, and tends therefore to have far more to it (even if we were to ignore the more obvious differences in the amount and depth of active mitigation, the responsiveness of mobs and the lack of absurd auto-attack range, etc., etc.). Members need to track each others' cooldowns, know or be able to predict when they'll be used and to what ends, know the priorities of a fight, and adapt per how they think others will. You'll see this also (at least, above the lower half) in Overwatch, where enmity (or "threat") doesn't rely on a banked number, but instead on a predictive situation and its seemingly viable possibilities. There, tanks are certainly distinct from DPS and Supports, but they're quite distinct from one another, too, and while their max throughput may be lower than DPS, it's precisely because that eHP actually matters (from not being infinitely sustainable) that they can leverage it to often outdamage DPS over a whole round. (And, the wonderful thing there, is that it doesn't make DPS any less valuable, as the dynamics of that damage are still so distinct, let alone the unique utilities provided by a given hero regardless of their role.)