Not exactly. While I agree that arm chair scientists should be more careful on the way they present their analysis and should have some background on how epidemic modeling actually works, I’m very far from being one (at least when it comes to this topic). I spent several years working professionally in epidemic modeling.
And yes, compartmental models are useful for many things. Which is why we should try to understand them and apply them properly instead of simply doing curve fitting or noisy and biased data.