This would have been a lot better with a “steelman” version of the “scientism credo” rather than the exaggerated form presented. I found it pretty alienating to try to read this, even though I probably agree with the thesis on the whole.
Yes, the case would be stronger with specific examples. However, I did not find it alienating, as examples of these 6 myths readily come to mind. We see people appeal to expertise all the time, rather than using their expertise to explain. There are lots of examples of people trying to "solve" economics problems rather than, as Thomas Sowell puts it, realizing that there are no solutions but only trade-offs.
As I understood the essay, Grothendieck's argument is precisely that the exaggerated form is what many believe. So I'm not sure a steelman would make sense here, given that his point is to argue against a specific set of widespread beliefs.
The steel-man version of "scientism" wouldn't be scientism, it would just be a straight-forward explanation of the scientific method and empiricism that acknowledges its limits.
A steel man version of scientism by definition has to be scientism, and I imagine that looks like strict determinisn/reductionism. I imagine it would go along the lines of "we acknowledge that our theories invoke notions of probability and since don't have a better solution we should assert them as universal truths."
That would be interesting! I'd love for someone to tackle that.