It would be good to the have the GroupBox title be “Edit RGB for IDn” so I know which ID I’m looking at, since the caption for the display box captures the mouseover, not the current selection. 
After looking at it through analyzing the RGB fluctuations on ID01-07, there’s no apparent set math nor trackable consistency here, at least in isolation. This is most notable against ID01 (chin), where the math sets (for 15%, 35%, 75% FAB vs default color) are:
LP/LK/HP: (-1,-2,- ; +2,-2,- ; +1, -2, -)
HK: ( -, -1, - ; -1, -2, -; -2, -2, - )
A1/A2: (-2, -2, x ; -2, -2, -; -2, -2, -)
, yet all are (4,2,0) initially.
If you want to analyze it as color strings, it seems like you could just reuse the existing math for the tint for LP/LK pale, HP red, HK tan, A1/A2 black, LP mech-purple, LK mech-blue, HP mech-red, HK mech-green, A1 mech-seablue, and A2 mech-dark-blue. But that leaves anything custom out in the cold. In that case, it might make sense to just reuse one of existing transitions:
Using more or less the LP/LK color transitions would be:
ID01: (-1,-2,-2;+2,-2,-2;+1,-2,-2)
ID02: (+1,-3,-2;+3,-7,-6;+5,-10,-12)
ID03: (+1,-4,-2;+2,-7,-5;+3,-10,-9)
ID04: (+1,-5,-3;+1,-6,-4;-2,-8,-6)
ID05: ( ,-5,-3;+1,-5,-3;-1,-6,-4)
ID06: (-1,-5,-2; ,-5,-2;-2,-4,-2)
ID07: (-3,-3,-1;-1,-3,-1;-2,-3,-1)
which starting with the dark-blue mech gives this transition for ID01-07:
def: (#(F000F16AF145F034F023F011F000)
15%: (#(F000F239F203F101F000F000F000)
35%: (#(F200F404F300F100F100F000F000)
75%: (#(F100F600F400F000F000F000F000)
or starting with a Spider-man colored gief:
def: (#(F800FF80FF60FF30FD00FB00F800)
15%: (#(F700FF50FF20FF00FD00FA00F500)
35%: (#(FA00FF10FF00FF00FE00FB00F700)
75%: (#(F900FF00FF00FD00FC00F900F600)
or with a Hulk gief:
def: (#(F050F6F0F0D0F0B0F090F070F050)
15%: (#(F000F7C0F190F160F040F020F020)
35%: (#(F230F980F260F150F140F020F020)
75%: (#(F130FB50F330F030F030F030F020)
I don’t know, that doesn’t kill me. I think that works OK for a starting point. It’d be cool to check against the known skin blends (the existing 10 skin shades) and auto-shade those if people are just copying those, but I’d think the basic LP/LK-esque math above should produce pretty livable results generally, as generally seen here.