How can tanking and healing be made more broadly useful in more "role-specific" ways?

[Shurrikhan]

I still think you can word this more intuitively, because I'm struggling to follow this.

In WoW, DPS utility mostly comes down to CD skills like Cloak of Shadows in obligatory soaking situations, Blessing of Freedom when not GCD-locked (which is why you almost never see it in the new Legion rotations, especially in situations ST-heavy enough to take Fires of Justice), Stampeding Roar, and maybe Touch of Karma, Die by the Sword / Enraged Regeneration, in the few cases where interception is possible, but this leaves a couple extra design issues I'd rather not have to deal with:

For one, their value is highly situational, and for that not to disrupt their balance, the unique opportunity has to come at equal opportunity cost to more standard opportunities or throughput (such as Feint coming at cost of Energy, or Aspect of the Turtle / Deterrence / Iceblock to uptime, or Defensive Stance to damage multipliers), which limits their usability and therefore contribution to identity. It ends up a pretty sticky issue from either a point of view of satisfying class play or encounter balance.

Moreover, anything short of those particulars of toolkit, or between their CDs, is then disregarded, rather than one person's CD being able to cover for the lack of another's — not just because he's swapping in to fill the need directly, which his toolkit may or may not allow, but because his toolkit allows the other player to fill it even without the security of his personal CD, etc.

I don't mind the idea of players having limited access to defensive skills, such as previously through Featherfoot, Keen Flurry, etc., but I don't want that to be the only times they feel like they can even interact. If they're made too strong compared to typical opportunities, or if the task is left too daunting or inefficient because it's tuned only around having those CDs at the ready, then the processes of party survival will remain tunnel visioned into just the tank and/or healer's purview. I'd rather these CDs merely be a way to really push those interactions—swapping in like half-tanks to cover chained blows that would otherwise capitalize upon the stagger dealt to the tank just before, or distracting the mob to draw it into traps, confusing it, enraging it, pacifying it, whatever—atop whatever means of interaction are provided through increased depth to and breadth of undermechanics (derivational enmity tables, mob focus system (allowing for enmity to be transferred or deferred through positioning), more than a single mob focus target (three heads = up to 3 targets), sub-targets (the legs of a mech, for instance), mob behavioral sets, the stagger system, etc.

So, rather than just having a taunt CD or a defensive CD on a DPS, I'd rather they have axis of control like timing and positioning where either of those, through undermechanics, can effect burst enmity-generation or increased personal, allied, or enemy-based (suppressive) mitigation. The icing is still good, too, so long as there's first a sufficient cake.

When I said buff I could have been more clear. I don't mean a literal buff visible on the bar. It could easily be a measure at which the enemy simply attacks/cast bar speeds up that we never see on the surface.

Likewise, I could have been more clear about what precedent or underlying hopes I was aiming for with my answer. A hidden effect that would seem to benefit the mob is perfectly fine. It could even be visible, though I'd rather it not be. But that should, imo, then be something very unique, fairly rare.

Essentially, I don't want these systems to just turn into something like enmity is right now—a sort of overheat mechanic—where you should go as close to hitting that enrage threshold without actually passing it (much like the tank being able to prioritize damage as much as still keeps him on top): instead, I want there always to be way to try to creatively exploit it. And these ways shouldn't need to rely on additional gimmicks; rather, simply the stats of the mob, one's team, and/or the conditions of the battlefield should be enough to say, with practice and a knack for the gamble, whether a particular strategy for manipulation should work.

If the mob is already really dangerous, then naturally you can't afford to send it into a berserk rage, whatever the conditions for that may be, and you may even need to hold off or assign someone with relevant utility to pacify it (at least until it catch on to what they're doing and enrages on them, specifically) because it will overwhelm you. But alternatively, a dangerous mob that you can kite while enraged such that it tires itself out (limited TP, as per 1.x design, albeit reversed) can actually be the safest option yet. Most of the time, neither may side may really come into play, but generally I want it to be a matter of the "mob is strong and irritable; don't make it use up too much at once, or it'll crush the tank" rather than "don't enrage the mob, because it has enrage bonuses".

While undermechanics are super important (and I'd be curious what line of work you are in IRL) you'll notice that I tend to focus on the "gameplay experience" rather than the underlying system design.

Pretty much everything I've come up with has come from an in-game thought of "wouldn't it be cool if we could do..." this or that, and it's only after looking carefully at the very simplest way to account for that huge collective of thoughts (those that have no conflict with each other or other parts of the existent game I enjoy or appear to be popularly enjoyed) that the system is developed.

(My line of work is day-jobs at this point, but since graduation has been centered mostly around teaching, tutoring, or acting as a research or instructional aide — speed-learning new things and then trying to teach them concisely, or developing study plans and providing more holistic insights. I realize my language here might not represent that well.)