I disagree - and it would be perhaps helpful to actually go outside and watch physical reality a little bit for tons of counter-examples. Yes, the gradient at t

I disagree - and it would be perhaps helpful to actually go outside and watch physical reality a little bit for tons of counter-examples. Yes, the gradient at the edge of shadows is washed out by the "flood light" of the sky, but it's still visible. This is very obvious with shadows from tall objects such as buildings or towers - just go to the edge of the shadow and observe.