Again, you're conflating a need by circumstance with a need for a circumstance. We currently need Lucid Dreaming. But we do not need mana to need a two-minute CD to be maintained. Nor do we need to limit mana management to, again, what can be done by an auto-clicker.
Is Diversion necessary in the way enmity is presently handled? Yes. But does it need to be? Do threat tables inherently limit their consequent gameplay to dependence on a cooldown ability to cope with that same system? No. Which is why nothing like Diversion has notably existed in other MMOs. It didn't even exist outside of Ranged jobs until Stormblood. It isn't necessary. Even if we don't want to sacrifice tank dps or return Ninja's once-notable erDPS via SW/SS, that's totally doable. You don't need to spend bloat buttons on it.
Melee-specific downtime, on the other hand? That is almost certainly a necessity of design when creating a slew of threats in a given encounter. Something which mitigates that, especially if it can be used skillfully, deserves to exist, because the thing it fixes is already about as fixed in itself as it's going to get. If the system has no inherent space for a control mechanism -- if further button count offers no real control by which to exert mastery and improve your situation -- then any skill there is bloat. Piercing Talon as it stands presently, is bloat. But the function it fills deserves to have something to fill it. Now, that doesn't need to be a ranged attack; in Heavensward, Monk's Meditate was the best melee downtime skill. But it makes sense when giving melee downtime, inherent to designing a fight that feels at all immersive, that you also give a ways to mitigate it. If preemptive movement truly feels like enough, then so be it -- the need is met. But if it's not, then something like Piercing Talon (perhaps instead in the form of a more fluid combo-connection into DS-ST or even Jump, SSD, and DFD giving ultrabrief damage and knockback immunity -- who cares as long as it feels good and it lets melee skillfully shrink the gap consequent to varied fight designs), improved as to be able to be used skillfully in the mitigation of melee-specific concerns in a way that synergizes with or compliments their kit, that has a very real place. Diversion does not. It's not just the lack of TP concerns now that gives Invigorate place. They don't augment gameplay. And that's what matters.
Consider: 20 more potency on Yaten-Enpi and it would be obligatory insofar as it doesn't cause desync, almost identically to Fracture among Heavensward Monks. Already, it can be a faint gain to use strategically at certain skill speeds, even without any benefit towards dodging. So why is Piercing Talon shit by comparison to EnhEnpi? And at present, Enpi adds something to the gameplay. Because of Yukikaze's flexibility if running mid-to-upper SkS, I can ensure I finish a combo just as melee downtime begins. And it makes it feel sharp. It feels rewarding. Because it hits for 280 potency less and has its job's rotation doesn't have a single GCD to spare unless taking painfully excessive Skill Speed.
Concept didn't kill Piercing Talons. Specifically designed internal mechanics -- the inability to cast PT without breaking melee combo and the sheer unalterable length of said melee combo, neither of which have any need to exist as they are (especially the first) -- did. The numbers did. Implementation did.
Enpi works. It works really well. But if precedence works as you're suggesting, that will be soon to follow into the trash bin, because it won't just be Piercing Talon. When you invite actual reduction of available options, rather than critiquing the ones that don't presently work, we lose actual stuff, not just the illusion of actions possible as your keyboard meanders fivefold to complete one decision. Alternatively, when you invite honesty as to how many decisions you actually have, then the focus is on their efficacy and on improving it rather than just taking anything below standard back behind the shed for the mercy stroke.
If I believed that there would be a replacement for the purpose Piercing Talon should fill, I would be fine with this. But all precedents are to the contrary; Monk's most central mechanic -- controlling the timing of their forms -- was gutted because someone decided "they have enough DoTs" without thinking about what those open-stance DoTs, ToD and Fracture (and Impulse Drive if not for Fracture's greater ppgcd) did for Monk. Fracture's purpose wasn't that it was a DoT. Touch of Death's purpose was scarcely that it was a DoT. Together they wove sub-rotations and allowed unparalleled freedom at higher Skill Speeds. But how will Piercing Talon be seen, now that it's failed to do what it should have -- as a ranged skill somehow there because... melee needed ranged skills? And so the trend goes forth -- melee shouldn't have ranged skills (duh), no matter how crisp and perfect they and other melee-downtime weaponskills can make mechanics feel when they're designed well.
Edit: We agree that PT needs a fix to stay. But we disagree on what PT actually means. PT to me is a way to make gameplay happen. That gameplay to me is not throwing a spear in a move called "Piercing Talon". That's its visuals, both the animation and its text. What does it actually let you think about and how does it pursue enjoyable gameplay within the job and among other jobs? That's what I care about. A PT off the GCD serves no more purpose than any of the several oGCD fluff-damage skills removed with Stormblood. It doesn't add gameplay unless, again, you conflate button count with gameplay.
For the... fifth(?) time now... I don't care about combo consolidation. I care about whether changes are made reductively or in the interest of a fulfilling product. Pretending that we have more going on than we actually do through the UI by which we interact with those decisions protects non-options at the cost of options, non-gameplay at the cost of potential gameplay. And I find it absolutely bewildering that you would trim anything that is currently poorly implemented as if it suffered from fundamental flaws.
You act as if combo consolidation will be some sort of dystopia of (weaponskills in numbers, abilities in letters) A3BC1DE1FG1H1122I222, but it feels like you'd be fine with losing applications of Doomspike in high-movement ST combat or of Elusive Jump and receiving absolutely nothing new to your gameplay for the next 3 years so long as you get to spend the most buttons possible on the fewest unique/separable actions so you can shuffle down those weaponskill keys. If so, that's pretty well my worst-case outcome.
This guy has to be trolling, right?
It's not hard to have Inner Beast and Fell Cleave, Steel Cyclone and Decimate each bound to different keys either. It's just damned useless. Would you like an extra Confirm button after your every Confirm button? No?! These entitled, lazy players these days...