The INVERSE of this principle is that one should not suppress or exclude phenomena which are known to exist merely because the explanation has no way to account for them. Adhering to this inverse is ethically just as compulsory, in the quest for understanding, as adhering to the original is in the quest for elegance and its concomitant convenience.
The molecular model of brain action does not have any way to account for the qualitative leap between stimuli and perception. Or, between perception and understanding. Sure, it can account for rods-cone stimulation, but where does that become perception? And how come some perception occurs with no such stimulaton apparent?