What makes a stat "interesting"

[Shurrikhan]

I see.

For me, any pure increase to power than does not change the relative power of your different skills as they are weighed against each other isn't even change, just scaling.
For me, any "improvement" is first a "change". Whether that "change" becomes an "improvement" depends simply on whether the majority of players think it has been changed for the better, where any players who feel it has been changed for the worse subtracts from that positive sum. The easiest way I find for a change to be considered an improvement is for it to offer more options, playstyles, and/or points of interest, rather than simply swapping in new ones for some or each of the old.

This could be something like, currently, enough Skill Speed making a BotD-drop rotation at a 2.33 GCD viable by adjusting the relative weight of DoT damage (one dropped Phb) and oGCD damage (maybe could have done one extra Geirskogul otherwise) against (Full Thrust combo's) direct damage. The old flow is still viable, even if stats aren't optomizing its play, but now another rotation can be used too. However, the way I'd want this to pan out is to allow that new rotation with minimal optomization loss to the main or "safe" playstyle. You'd still be gearing for the new playstyle of your own, but you wouldn't be losing out so badly on stat weight for having done so and then using the original combo anyways when its use might be more appropriate. More options. That much can come from the nature of stats. Of course, it can also come from talent customization, adjustment to the original rotations, and so forth. What matters to me is that the maths behind a job's skillset and how they interact with the wider game rules allow for lateral complexity — a multiplicity of choices, with situational, rather than niche, use, and an incredibly high skill ceiling to ever optomize back into a singular rotation, should one dare try.

It's this matter of change vs. improvement, aimed at maximizing options or points of interest, that makes me wary of suggesting certain quality of life changes to our skill sets, even.
For instance:
- River of Blood (ARC 48 Trait) changed from "Your Venomous Bite and Windbite have a 50% chance on critical tick damage to reset the cooldown on Bloodletter and Rain of Death,"

:: You should track the 3-second tick and be sure to use Bloodletter in time to have the ability consumed when the ticks go off.
:: This is prioritized further relative to other uses for the given time or window (other oGCDs) as your chance of a refresh continues (number of DoTs, whether they were crit-embonused on application, atop your base crit).
:: As refresh chance increases but the Bloodletter can nonetheless be triggered before the 3-second global tick, Bloodletter is deprioritized in order compared to advancing the CDs of certain other oGCDs if improving windowed coverage. (This part especially is risky, meaning a higher skill-ceiling with greater possible error.)
:: Opportunities to weave a Bloodletter perfectly increase with open (or oGCD-weavable) time per GCD. (E.g. Straighter Shot or Feint in WM)
:: A Bloodletter within a half-clip Emyreal Arrow means that the next tick will go off in less than 70% a GCD's time, roughly cast + .2 seconds. It is unlikely you will be able to trigger the Bloodletter in time. It may be worth clipping further to avoid waste.

- to "Critical periodic damage dealt by Venomous Bite or Windbite advance the cooldown on Bloodletter and Rain of Death by 3 seconds."

At a 30% base crit, 40% with Straight Shot, you have 6 ticks of a total of 40% chance using both DoTs on a single target to refresh Bloodletter. (Theoretical average 2.4 refreshes per 18s)
At a 30% base crit, 40% with Straight Shot, you have 12 ticks of a 40% chance to advance Bloodletter. (Theoretical average of 14.4 seconds advanced, or 32.4 seconds of Bloodletter recharge per 18 seconds (2.16 refreshes))

>> This would require some retuning, obviously, but the concept is workable. In the end, balance will also have to consider whether greater reliability or greater potential burst should have higher theoretical average output.
:: At any greater than 4 seconds time remaining in single-target, there is no chance of a wasted Bloodletter or Bloodletter charge. This makes it easier to track Bloodletter priority, slightly smoothing reuse of Bloodletter during an opener, especially if in WM.

This may be considered smoother to any number of players, as there are now two ways to check your risk and you have an easier estimate of your total uses. However, it is not necessarily more challenging or interesting. Heck, this version, unlike "Your Venomous Bite and Windbite periodic damage ticks reduces the remaining cooldown on Bloodletter and Rain of Death by .75 seconds, and their critical ticks by 1.5 seconds" or whatnot doesn't even adjust the critical strike weight except in cases where immense critical strike chance and its consequent Bloodletter refreshes could force the Bard to desync its oGCDs from its burst window. As such, I would guess it would be unlikely to be found an "improvement", depending on how much most Bards like interacting with the prior options and points of interest, which the change has reduced in number, rather than expanded.