[Autumn]

The stat investment for such would be tremendous and heavily unreliable against to-hit builds, making to-hit actually have a purpose, and as stated in the original post, the numbers to actually determine damage reduction could be tweaked.

I'm pretty sure when you hard boil down to actual statistics this is very much currently untrue, you are using one frame of reference without the evade buffs in question I already mentioned, there are some really really chaotic evade buffs presently in the game that make even hit builds cry quite a lot when they are active, disengage + Miragewalk, chaotic form, Unarmored Prowess + Hare Stance, Night Shade etc.

This would get chaotic very fast, I don't think the goal of improving evade should be inflating evasion to such an egregious degree and dismissing it as "Oh we'll just fix it later."

I think how you currently improve evade is not by buffing it at all, I think you just make basic attacks seem more worthwhile as an option over the likes of autohits, probably by reworking some melee weapons to scale from Guile instead of Strength or something similar. Its not that basic attackers don't do enough damage its more like they don't have stat efficiency like autohitter tanks do.

Currently in the Evade vs Hit matchup, Evade tends to win currently given buffs like above, making certain classes more evasive instead of tanky allows that class to flourish with damage buffs, the likes of Flottement could see a rise in power again, or absolute pace being reworked to not actually be bad in almost any given scenario.