[One Winged Angel]

Score is likely far from optimal, especially if the same formula is being used to determine both multipliers. Best I could surmise is that some nonlinear regression was created for this and both misses and combo are being used as inputs for determining the multiplier. Given it's a regression, probably better to boost raw score slightly more by missing less compared to what I did (combo nearly half the song and try for ~50% misses on what's remaining - maybe go for 30-35% misses instead)

edit: though after seeing the multipliers, it's clear there's different formulas (this post was an edit I put into my previous post before I saw what the multipliers I got were, but moved it here after there was a response for visibility)