Should color equality consider missing channels to be unequal to 0?

[nex3]

Currently, the color spec defines equality as just each component being fuzzy-equal to one another. This implies that lch(50% 50% none) == lch(50% 50% 0deg), since we say missing channels are treated as 0 anywhere they aren't mentioned explicitly. Is that desirable behavior? Or should we have a special carve-out for missing channels that makes them only equal to other missing channels?

[mirisuzanne]

I think we likely want a carve-out for that, since we've gone with a more strict definition of equality.

[nex3]

We could potentially add (in a future change) a color.visual-equals() function which returns true if two colors will look the same on all displays (essentially, if they're equal in XYZ space).