I make this call based off how the fight progresses. Regularly, there is not much use in using Swiftcast for other abilities specifically. However, if we were talking about maintaining the 3 main DoTs solely, I would think that to fully flesh out the limits of the DoT timers in relation to the server tick, using Swiftcast to apply DoTs regularly is in fact good. More to the point if a DoT is falling off (3 seconds or so) in a situation where movement is currently being forced. I roughly find that if a mechanic is happening, it does not last more then 4-5 seconds, in regards to you positioning yourself in a safe area.
Now often for myself in T13, I find that Miasma I is roughly at like 5 seconds when the Earthshaker marks are disappearing. If I am selected, by the time the earth shakers finish and I'm safely able to cast, Miasma has fallen off for about 1 - 3 seconds before it is reapplied, this mind you is with a Fester at the ready. To get around this, I simply refresh it with stutter step as well, right before the earthshaker hits. This will save me at the least 1 server tick of the DoT but depending on the timing, possible two from preventing it from falling off.
Using this same instance,let's make a controlled enviorment. Let's say you are in charge of taking down the Shadow of Meracydia alongside another DPS. Shadowflare needs to be replenished when Megaflare is casting as well as recasted again at a different location and the boss is constantly being moved away from the Meracydia into these new locations. I would argue that for the sake of the overall situation, Hardcasting those two ShadowFlares, and using Bio II -> Bio -> Swiftcast -> Miasma to burst down the Shadow is more beneficial to what is happening. I wouldn't be able to easily tell you if this was a DPS gain or loss though.
Eventually, to find out on a fixated target, we'll have to look at shadowflare base 3 second cast time, stack that by a certain amount of duration and then compare hardcasting a Shadowflare in comparison to a DoT tick we are saving within the same duration. Since ShadowFlare needs to be replenished every 30 seconds, we can assume that after 1 min you have in fact lost 1 second worth of GCD due to its cast time.
3mins -> 3 seconds of GCD -> 1.2 GCD
5 mins -> 5 seconds of GCD -> 2 GCD
10 mins -> 10 seconds of GCD - > 4 GCD
13mins -> 13 seconds of GCD -> 5.2 GCD
Balanced by Ruins,
3 mins -> 80 Potency
5 mins -> 160 Potency
10 mins -> 320 Potency
13 mins -> 400 Potency
Let's consider Miasma I is solely used with Swiftcast to reapply and we were saving 1 DoT tick by allowing it to reach it's lowest point of 0 seconds and then reapplied immediately. This would be 1 DoT tick (35) times number of minutes.
3 mins -> 105 potency
5 mins -> 175 potency
10 mins -> 350 potency
13 mins -> 455 potency
If these numbers have weight, assuming a Perfect World Scenario, the increase from saving a DoT and theoretically gaining a server tick via a Swiftcast will beat using a Swiftcast with Shadowflare if my math is correct.
However, it is a perfect world scenario on a standalone target and assumes the server tick will always occur the moment you refresh your DoT as soon as its falling off, to get a confirmed extra tick and not including any outside factors like mechanics or adds.
Places on average where you cannot cast a Shadowflare reliably for a decent amount of time, use it to apply DoTs. Example: T11 Add phase, and attempting to DoT up the Egg node specifically. Due to the nature of hitting the egg node, and its movement around the arena, we know that you can either get LoS'd by it or it will move out of melee range of you and is now invincible, so Swiftcast to apply DoTs here is also great. This doesn't break the chain of command in casting ShadowFlare either, as the next ShadowFlare right as soon as the add phase ends should be hardcasted.