I'm not following you. Your examples are of the inherent risk in the activity or device - a person drinks coffee regularly, and then burns herself after drinking safely for years - this is just a risk of a hot drink, it could happen at any time. Having a sign on the item doesn't impact the risk for a regular user: the regular user knows the risks, and will ignore signs they see all the time.
Signs might protect a new user - 'this is a guillotine if you hit the button it will cut your hand off'. Is that the point your making?
Signs might protect a new user - 'this is a guillotine if you hit the button it will cut your hand off'. Is that the point your making?