I have to say, I'm not a fan of proficiency award systems, if only because they inevitably tell you how you should play, punishing the player for any unconventional or more extreme playstyles. Often the very things that can make a composition interesting, systems like this can tear down. Put simply, they are too often wrong, and much too often stale.

In other ways, and I mean this much neutrally, it can directly conflict with the similar methods we have of similar fudging the LB. If you ever stacked for extra damage taken on a Bahamut's T5 Plummets, or had your Black Mage run up to the tank with Manawall to take would-be critical damage from T9 Nael's Dragonfire Dive, you should know how these little bits likewise have a nostalgia of their own. You would now be likely be punished for such gimmicks, just as you might for Holmgang-ing a tank-buster rather than mitigating it, etc. Even if the system could determine who all dodged a given AoE (checking players in the targeted zone originally vs. those hit by it), how is it to award you for spreading out as to reduce raid movement necessary when a single player is targeted by one? How is it to award you for finding the proper balance the number of people involved in splitting an AoE and the amount of people (those hit) affected by the AoE's debuff (which cannot be split)? While designing a system that can do all that may be theoretically possible, it also seems absolutely unnecessary. Let the mechanics, balance, the players, and success itself be the judge.

I don't think there's any need necessarily for an Enchained system, but there is certainly clash with the rotational meta. For one, most classes only have 1-4 actual weaponskill or offensive Spell options in a given moment. Any combo class has no choice of ability mid-combo. A Monk can only choose from among its choices in stance (Dragon Kick, Bootshine, or Arm of the Destroyer; Twin Snakes, True Strike, or One I'lm Punch; etc.). A Black Mage isn't likely to end its AF burn phase early just to initiate or join an Ice chain. Numbers given, there's little point in reapplying a DoT early just to apply a particular element. Elemental choices are already highly limited, highly concentrated. And most weaponskills, frankly, don't seem at all related to any sort of element.

To continue with the Black Mage example, those original rotation confines of AF and UI strings mean that your Black Mages are either going to need to maintain a particular rhythm together (which could be a fun bit of gameplay, if not for how few the internal options are by which to usefully compromise phase timings) and other casters are probably going to have to match themselves to their pace. WHMs could hold off briefly on their DoTs in order to stoke the fires with their wind magics, punctuating the burn phase with a magic burst (again, not bad when envisioned in glimpse). Worse perhaps, you have the utterly unaspected Arcanist and its jobs. That could work to your advantage, such as by being the one type who can boost any element's chain (but can burst none, except perhaps by overfilling it or whatever spitball extra concept one might imagine). Not necessarily bad, but potentially very different, but where the best way to use the new system is probably to try to make it as similar as possible to the way things were already done.

The harder part will be the physical classes, especially the combo users. oGCDs may be necessary, allowing a supporting player to time an initiator to the start of another's combo. Alternatively, combos themselves could be considered some sort of unaspected weaponskill boost, meaning that the only real controllers of the skillchain would be the Bards' choice of Heavy Shot vs. Straight Shot, the Monks' choice of DK vs. BS, Twin vs. True vs. 1lm, and so forth. But that hardly sounds compelling.

That said, the last thing I feel an additional mechanic like this needs is to swap between metas like some sort of quick-time event. (Especially when the mechanic itself might not seem to do any more than that... See below.)

The Skill Chain Rules provide a solid enough compromise in that there is seemingly only one initiator, one finisher, and one burst, a maximum of three steps. Any number players being able to react within the 6 seconds means that no one can take the better use of the initiated skill chain, but it also means that, resources willing, everyone in the party should be making use of every initiator. And with 6 seconds, two and a half to three Global Cool Downs, there is very likely time to do so. With only one progression path and only progression skills counted, no one can potentially ruin a chain by typal error, either.

However, there are two negatives on the opposite face. The first is that this means you can only use a chain every 12 seconds, or slower, because of the guaranteed 6 seconds' skillchain and 6 seconds' burst time. (Or can one of each opportunity be up simultaneously? If so, then per 6 seconds.) I suspect that's a non-issue to you, given that you were thinking of adding a limiter to the systems use anyways via Enchained?

The larger problem is that this keeps the system shallow. With only one progression direction, or only one possible "answer", there's no element of choice. Moreover, with those 6 seconds, there's almost no way that the whole party to will fail to answer in time, regardless of how poor coordination may be. They could likely turn off the skill-chain UI and keep on chaining regardless. If not, the meta would probably quickly accept only those chains for which that could be done unless intentionally crafted otherwise. Gameplay is almost irrelevant to how the system pans out. You might as well have a chance on hit to deal bonus damage.

P.S. I'll try to give a counter-proposal that might approach what you're looking for without hitting these issues when I can. That may have to wait until late tomorrow though.

First, if all you're really looking for are more reasons to differentiate your rotations based on other players, that much can be done through adjustments to rotations and their skills, without any need for an new system. What will likely appear to be done over all rotations is to slight de-optomize the optimal rotation, making other (albeit equally purposed and cohesive) rotations less inferior; in actuality it will be exact other rotations brought to the fore. Simply put, it will be intended emergence; you're trying to adjust pay-offs and rotations such that composition can have an effect. Effects shared between multiple classes are a good start for this. Some of the worst for this are skills with differing priorities or varying in raid-shared to personal effects that are necessarily linked (Disembowel and Chaos Thrust); as long as Chaos Thrust retains such massive damage, this makes it very difficult for one Dragoon's Disembowel to allow any other a different rotation; they must go through Disembowel, themselves, to reach Chaos Thrust, which they can scarcely afford to let drop, anyways. Consider how many of these jobs with such rotations or gameplay that are problematic for compositional variance have problems perceived in the same skills for different reasons. (For instance, could changes to Whirling Thrust and Fang & Claw, making them less identical and/or less rigidly timed, fix these both the compositional-independence and issues more widely perceived, without creating new issues?)

With enough such changes, you may well see that jobs individually better appeal to their players, while players are each much more interested in what others of the same job or with any shared effects are doing. This won't have the same broad effects as a universal system, but it may well have a greater burden of knowledge and carry a greater feeling of needing to know the playstyle of your partnered characters; the fact that this direct, potentially rotation-affecting bond exists only with certain classes may also be perceived as deepened identity.

But if that's not enough, and you really do want that new system to shake things up.

Let me append three major conditions.

1. There must be reasonable excuse and impetus for the system's addition. Otherwise, it feels like an admitted mistake, rather than added content (in the way of systemic gameplay). This can be a theme for a new expansion, for instance, or a new (rather than a retake in) announced gameplay direction, etc.

2. If a system that would adjust gameplay is to be used universally, then the changes to universal gameplay to support it must in themselves feel beneficial and progressive.

3. If the system is to be used tactically (in the sense of choosing to save opportunities for greater opportunities), it will need a limiter; this can be through windows, a resource, limited triggering events, conflict with general rotations, or whatever else.