A critique of the classic layered model.

Author says that this is better:

I'm not convinced.