Off the cuff I am thinking that stability requires stabilizing forces that come from somewhere and that that somewhere takes energy. In an abstract model which causes more drag to roll 30 degrees, a wing with dihedral or a straight wing?
I'm not sure it's the same thing as unstable but making a wing that stalls inboard first has got to be more draggy than one that flys the same along its whole length.