Click-bait titles can be moderated and edited more easily.
Especially algorithmically, we're a way off being able to do the same with images.
Except when the aggregators guidelines specify that titles must remain verbatim.
People sometimes think that about HN but of course it's not true.
Good point. I wonder if this might be measurably worse. A click-bait image might have the power to grab your attention from your peripheral vision in a way that pure text, however click-baity, can't.