Because you don't have to use the debugger or mentally map the code all the time to see the (actual) shape of the objects at any point in the call chain?