This hatchet job on academia is insane. Some of the points are well-made, but largely it's a denial of how complex systems including the weather are routinely simulated. Code that's been validated is often left in its original state rather than refactored. https://lockdownsceptics.org/code-review-of-fergusons-model/

I'm all in favour of funding and staffing a full refactoring, but it is NOT that unusual for scientists to choose ensemble methods and replicate populations of simulations results rather than individual ones. We got that habit from experimental science, which has no other choice.