Thank you for the detailed report. I have actually been keeping an eye on accuracy for a little while now (since before FBQ), and while as I said I have not been able to identify any issues with number generation that I can clearly separate from cognitive bias, I do agree that it doesn't "feel" right. I do have some ideas that I am going to try out to see if they help.
One thing that does contribute to this is that accuracy is capped at 95% for any given attack, so this can cause e.g. Command + Overpower still missing even a Scavenger on occasion, which makes the player think "how did that miss?!" I think I will remove this cap, at least for players, since it isn't really benefiting the gameplay.
The subway enemies are intentionally quite weak since they're just tutorial fodder and not meant as a serious obstacle (before I toned them down, I used to get complaints about them being too hard for the tutorial, in fact), but I am looking into bringing down the difficulty spike when entering the hospital just a tad.
Most AoE debuffs are intended to have a lower success rate than single-target ones, but I think this isn't configured properly in all cases. I'll do a pass over these. Alluring Scent in particular is probably too effective, especially when facing multiple Conservators.
I'll run some numbers on Champion and see if it matches up with where I want it, but I will also consider the suggestion to have it protect the back row specifically. Targeting is weighted by position, with Champion adding a big number to the user's weight, but it doesn't make the user a total magnet (though one might argue that it should).
I am not sure I agree that Dispatch should only cost HP on a hit, since it is a fair bit stronger than e.g. Flashing Blade and the real cost of using it is split between HP and MP, but I do see your point. I will think about it.