A simulation is an experiment, and a MC simulator may be considered a machine that can churn out a series of experiments. The simulator will behave in a certain fashion (i.e. produce symmetric, asymmetric, normal and skewed, with thin tails or long fat tails) depending on the tool used to build the machine (i.e. the choice of distribution). By definition it will always be inaccurate and an approximation to the real world.
The Monte Carlo Simulation resource page has moved. Check out the new home for the Monte Carlo Simulation page.