You have a thesis which is that people place too much faith in models and interpretations thereof, and thereby cause harm.

The way you argue for your thesis is to make assertions without evidence. This is not compelling.

I don't think it's likely that there has been adequate research into the extent to which over-reliance on models is a problem in general, so we're probably stuck with case studies for evidential support.

Thus, I am looking for at least one worked-out case study. If not in this article, at least a reference to one. The not-worked-out examples include Covid epidemiological modeling where, at the level of generality that you've stated it, it is contested whether following modeling led to any serious problems or only benefits (as opposed to any available alternative)--but you don't engage in the contest and instead simply state it as if it's a known and accepted fact; and climate change where your observation about models if used to discredit their accuracy is not distinguishable from the maelstrom of crackpottery around climate change denial. Even in the case of sensational climate-change-caused-disaster reporting, you don't manage to give a single example of the type of overreach you think there is and draw the link to modeling instead of hype-sells-so-generate-hype-from-everything.