The problem is, that the state transition information is usually a complex set of events unleashed upon the little world build into objects. Its basically a small reality simulation, parametrized by all sides, similar to a physics simulation with the external world filtered being input.

Now, if we said to someone to model a weather prediction via state-machine- that would be obvious madness. But if we take a small object (like a cubic meter of air) and modelled that part to handle inputs and transfer forces, that generic statemachine will do the job- because the atmospheric ocean of statemachines knows more about the whole system, than a single statemachine.

My point is- there is state in a system, that is not explicitly modeled.