Imagine a Raidwide attack is coming, and you want to put up a barrier against it to protect the party.
Current SGE:
You press Eukrasian Prognosis. It costs 800MP to cast, and upon breaking, grants an Addersting. This can be spent on Toxikon for 370p damage.
The 2 GCDs you spent give a total damage of 370p, whereas spending those 2 GCDs on Dosis would result in 740p, for an overall damage loss of 370p per barrier you cast.
My design for SGE:
You press Eukrasian Prognosis. It costs 500mp to use Eukrasia, and 0mp for the GCD itself (meaning that if you're out of MP entirely, you can still spam Prognosis heals instead of just standing there).
Eukrasia's activation generates 5 Toxikosis, a gauge that replaces Addersting. 50 Toxikosis can be spent on a burst phase to deal big damage via faster attack speed (which results in Kardia being triggered more often).
Upon the barrier breaking, you (the SGE) gain 'Phlegmatic', a new buff, and everyone in the party (including you) has their barrier buff replaced with 'Second Opinion', a new buff.
'Second Opinion' can be consumed with Pepsis (5s CD, costs 500MP, generates 5 Toxikosis) to trigger a healing effect, just like when it consumes a barrier, allowing for a more 'reactive' kind of gameplay
'Phlegmatic' allows for the execution of Phlegma, without consuming a charge. Instead of Toxikon being the 'refund' attack, it'd be Phlegma which hits harder.
Due to potency rebalancing, Dosis would be 310p and Phlegma 500p, so the damage lost by using Barrier>Phlegma instead of Dosis>Dosis would be just 120p, not the current 370p.
My design for SGE (but the player is a tryhard):
You do not press Eukrasian Prognosis at all. You know the raidwide is coming, and prepare for it in advance.
Soteria increases the healing of Kardia by 50% of the damage of the attack that triggered it. This means that a Phlegma would cause more healing via Kardia than a Dosis, for example.
Zoe adds a barrier equal to 25% of the healing dealt by Kardia, stacking with itself to empower the barrier. This allows the player to protect the team without having to let up on the offense.
Krasis reduces the cast time of the next 4 spells with cast times, by 3 seconds. This opens up doubleweave spaces, and even allows for faster resurrection casts if required.
Pankardia (new) causes Kardia's healing (and barrier) to be applied in an AOE, but at 50% reduced effectiveness for all targets that are not the main Kardia target.
All of these effects have 4 stacks (as a reference to the four humours). Consuming a stack of any of these grants 4 Toxikosis, and triggering Kardia grants 1 Toxikosis. Thus, each use of these augments grants a total of 20 Toxikosis overall.
Philosophia instantly grants you 4 stacks of all 4 augments, and increases the potency of your next 4 heals (or next 4 Kardia triggers) by 20%. Its CD is also reduced from 180s to 120s
As such, because you are a tryhard and know the raidwide is coming, you use your Kardia augments to apply a Zoe barrier to the party in advance, fulfilling the 'protect team' part of your role without having to sacrifice any damage output.
Furthermore, after the Zoe barrier is consumed, it is replaced with the Second Opinion buff, so you can heal the party after the raidwide with Pepsis.
If you tried to learn the tryhard methods of this design, but felt overwhelmed, or you were to panic and think 'I don't have time to put a Zoe barrier out, the party's gonna die aaaaa', you would be able to simply fall back on a Eukrasian Prognosis barrier. The potency loss (compared to current game) is far lower, the MP cost is lower (so it's far more likely that you will have enough even if you're running on fumes), all of your basic GCDs would cost 0MP (so you are always able to cast 'something' instead of standing there waiting for MP ticks/Lucid). The majority of the 'optimizations' you could do with such a design would be entirely optional for 95% of the game's content.