I do not have a fluid dynamics background, but my hunch is that at the speeds involved, and the fact that there is a column of air from the cannon already surrounding the projectile, that this makes this simulation close enough.