Modeling is using data inputs, independent and dependent variables, in order to answer a well-defined question of which you do not know the answer and generally these are fluid and change over time. It is not having a hypothesis you are trying to prove (scientific method), which I think is the stance that people get upset about. It implies manipulating (testing, experimentation-- scientific method), which would discredit and disqualify any model. Further all models are obviously erroneous-- sure you can look at p-values and r-squared and everything else used to account for the error (margin of error acceptable), but like I said these are more like how we look at polls than experiments.
Your response just seemed lazy and sloppy. Modeling, in any understood way as the term "modeling" means in 2020, is not scientific method. In a very fundamental way, you are wrong. I think your anecdote about his treatment of servers is a better way to attack the man then his having some beef with the models.
Also Jimmy, I apologize if I offended you and being presumptive. Not my intent and I can see where I might have been disparaging after re-reading, but I was mostly just giving you an elbow in the ribs with a smile.