RDM Scortch

[Archwizard]

I would see them as a situationally useful tool to make use of our melee combo when either the boss (via mechanics) or the rest of our party (via buffs) or the game itself (via procs) makes it useful or necessary to go for it now, much like the real-world addition of Reprise. Using Reprise is a potency loss too, so in an ideal situation you'd never press that button at all, but if it lets you land your melee combo in a Trick window without wasting mana or keep uptime when the mechanics demand you move continuously for a long time, then it was worth pressing, because the smaller loss led to a bigger gain.

That's an apples-to-oranges comparison. Reprise can be used to prevent from overcapping Mana in a multitude of situations (such as spending Mana during mechanics where entering melee is ill advised, since we can still build Mana at range), and comes at a higher potency-per-Mana than any melee skill in our combo, meaning there are ideal situations where it can intuitively come as a substantial damage gain (i.e. if you're near cap and rolling in Verquick procs that would be overwritten by your set-up Verfinisher meaning lost Mana anyway -- one glance, "I just need to burn this proc", E.Reprise, burn proc, follow with the opposite longcast, now you've instantly made back the cost within a negligible difference and turned what would've been lost Mana into damage, hit your melee combo with a clear conscience), acting as another tool to help us control RNG.
Not to mention that its use as our Scathe/Ruin II attempts to fill an open mechanical niche in our kit that we have nothing else to provide for except stutter-stepping and Swiftcast.

Meanwhile, you've listed no cases for Verflare/Verholy where your "better" Verfinisher wouldn't still outshine them if it was available, nor cases where they are made more "necessary" by encounter mechanics, as they would all attempt to fulfill exactly the same niche in the kit and position in the rotation. You're still relying on players to be able to substantially math out in their head in a split second whether to take the loss or not, which without using Acceleration every time would mean praying for good RNG, which in itself would determine the value of said loss within the situation.
In any realistic scenario it's overly-complicated, impractical, and goes against the "at a glance" intent of our design philosophy, all to satisfy an obsessive need for purely aesthetic "balance".

OP, I apologize profusely for any thread derailment that may have occurred by me attempting to explain this same concept half a dozen times. Singularity, please drop it.