Astral Fire/Umbral Ice timers are actually bad mechanics at a meta level.

[Grimoire-M]

I have to somewhat agree with Rongway's conclusion that Fire IV would need something else gating it to make this work. My issue is more with the framing of AF/UI as being bad for having the tension they do. It would be difficult to maintain the same level of forethought that the timers do encourage you to have once you start accounting for how some of your mobility options come into direct conflict with them. Particularly when you factor in that it also rewards you for gearing into spell speed and using Leylines in particular. My overall interpretation of the job is that the reason it's a successful turret mage is because the timers allow for some level of flexibility while still putting some rails on how you handle things, and Polyglot and Fire IV both were ways to reward you for maintaining those timers (Fire IV more for Enochian, specifically in regards to HW). And then you get into edge cases. Manafont doesn't work unless you have a timer. Firestarter and Fire I/Paradox have no explicit purpose without that timer. Some of the value of Triple/Swiftcast is lost without those timers. It adds tension, sure, but not excessively so, because it doesn't ask that you juggle a bunch of secondary resources. Astral Fire and Umbral Ice are the core, including the timers themselves. At some point you have to ask, "Does taking this away add anything?" And judging by the hoops Rongway jumped through to get it to work without it, I'd say no, which is distinctly different from how Monk was handled previously. And honestly new Summoner gets punished more than Black Mage ever has by downtime, which isn't something I ever thought I'd say given where both were back in 2.0.

[Rongway]

At some point you have to ask, "Does taking this away add anything?" And judging by the hoops Rongway jumped through to get it to work without it, I'd say no

I think one of Taranok's major points though is that making UI and AF persist until zone/death would change nothing for the top 15% BLM players, but BLMs in the lower percentiles would have a smoother time, enough that they're not severely punished for being half a second too late on a midphase Fire cast, but the lower percentile players would still have enough other aspects of the job (continuous GCDs, preemptive choreography, the MP dance) to improve on.

I must have done a bad job of explaining my interpretation of the suggested mechanical changes if it sounded like I was jumping through hoops? Applying edit to previous post.

[Grimoire-M]

I think one of Taranok's major points though is that making UI and AF persist until zone/death would change nothing for the top 15% BLM players, but BLMs in the lower percentiles would have a smoother time, enough that they're not severely punished for being half a second too late on a midphase Fire cast, but the lower percentile players would still have enough other aspects of the job (continuous GCDs, preemptive choreography, the MP dance) to improve on.

I must have done a bad job of explaining my interpretation of the suggested mechanical changes if it sounded like I was jumping through hoops? Applying edit to previous post.

I understood your intention well enough. The 'jumping through hoops' comment was only referring to the amount of programming that would be necessary to facilitate the change in the first place, while also attempting to showcase how the timers limiting BLM also encourage it to think about its sequencing, and highlight that aspect of the job during the rotation, even if it feels rote at times. They're intentionally just loose enough to work with, and no more than that. Removing them essentially means you have to account for new edge cases where there wasn't any before and add more mechanics, all to solve a pain point that doesn't need as much attention as it seems in the first place, given the OP themself admits it's just about getting access to the core rotation. That's the point. The timer facilitates that with less text and cleaner intention. That said, regarding your particular choice of replacement for the timer, I'd probably make manafont and fire III give the new stacks for Fire IV if used in AF and UI respectively, rather than create this awkward situation where you have floating stacks of it that you have to juggle through multiple AF windows or consume in some manner. Less abilities would need to be changed while still encapsulating the original intention, with the only weird additional case being able to move Paradox/Fire I anywhere within the AF window, or skip Fire I entirely unless you need the Firestarter proc. Though even that could be solved by giving Paradox & Fire I a shorter cast time so that there's a smidge of extra movement for using it.