I'd expect that even if he'd been totally blind, given the normal process of mixing paints on the palette before painting with them and that this is imprecise.

(Obviously I don't think he was totally blind, I'm only saying subtitles of colour aren't evidence of anything).

He would not had subtle shade play in all those colors. You would expect shades to be used subtly badly.

That's exactly what the researcher found -- excessive green where it didn't belong, but brown did.