Old databases -- has no activity
Doesn't mean they can't be corrupted due to an IO subsystem glitch. It you absolutely don't care about them at all, fine. If you use them at all, they should still get consistency checks every now and again.
I'd have to agree, and say that I would either schedule regular CHECKDB runs, or drop the databases entirely. If the data doesn't matter, then drop it. If you can't drop it, then it does matter, and you should run CHECKDB regularly to spot errors.
And yes, I've seen one system where a hard drive had failed silently, and no errors appeared. Then another drive failed catastrophically, and all the errors came to light (RAID5) as actually corrupt data.