Used it a couple of times when n was inherently very low. And while I've never hit a case I would think it would have merit with data known to be pretty close

Used it a couple of times when n was inherently very low. And while I've never hit a case I would think it would have merit with data known to be pretty close to properly sorted.