I think there's a lot of good to come from studying things like Alcubierre/White drives. They are unlikely to actually work, but that doesn't actually matter at all. When we understand WHY they don't work, we'll be that much closer to coming up with something that does.