Assume: We have a box that reliably detects evil.
Assume: Demons do not exist.
Scenario: You are walking outside, in the middle of a major city, with nothing but people around you, going about their day. Your detector is mostly in the Good zone, however, throughout the day, the needle fluctuates towards Evil, and once even pegged all the way to the other side.
What could you (tentatively, of course) conclude from this experiment?
That evil passed by. Mostly it would be used to validate suspected evil situations before further action was taken.
The relevance is that art helps you see something. It doesn't get anymore relevant than that. The catch is that it is that there has to be an actual experience behind it or the art is just empty.