Simulations can certainly be a way of doing science. Of course you have to ask how realistic the simulation is, but that's true of any model -- you always have to be aware of your simplifying assumptions.