Monte Carlo simulations are fine if you have a good understanding of the inputs, the range of values, the distribution of those values and how each input may interact with the others
It doesn't give you a "right answer" but only a "most likely " distribution
It is a lethal technique if there are "Black Swan" event out there, or you don't understand the inputs and their interaction
Reading Engine's excellent post I'd say it was fine for canopy thickness issues but something to avoid for simualting opposition tactics etc.............