1. It does not produce repeatable results given the same input data, due to bugs. "Averaging" the results is suggested.
2. The algorithm uses equations of unknown purpose and origin.
3. The source code is hidden from independent view.
Passing off the results of this model as "science" is risible.
Unfortunately, that's been true for 90% plus of all the science to date on this pandemic :(
Reading some of these papers and thinking of how merciless I'd be if I were their peer reviewers... then seeing people treat them as absolute truth... it's depressing as hell.
(I was one of the exceptions, and man, was that ever frustrating.)
All code is trash, but not all code is trash for all the same reasons.
I think the article author mentioned that just to show (s)he understands stochastic modelling. As can be seen from your comments and others, it's apparent some people/scientists either aren't reading the article properly at all, or have a massively wrong understanding of what Monte Carlo/stochastic modelling techniques really are.