What I take from that was the interesting point that you can take a principled approach about what "correct colour perception" means. But there's no such stable

What I take from that was the interesting point that you can take a principled approach about what "correct colour perception" means. But there's no such stable basis for deciding that with olfaction given that it seems like an ad-hoc jumble of molecules to sense and there's such a wide and, seemingly random, variety of gene variants in the gene pool.