Concretizing the abstract is an interesting blog post in that it makes a relatively cogent argument for non-reductionism
Quite the opposite.
When high-level abstractions fail to accurately reflect the actions of high-level objects, this is not in any way a refutation of reductionism. Rather, it is exactly what we would expect if reductionism is correct. If you try to model a billiard ball as a single object, rather than a collection of quarks and electrons, of course it will display behavior that doesn’t fit your model. It is accuracy of high-level models that challenges reductionism, not inaccuracy. If you could model the billiard ball (or a human being) as a single object completely and consistently without any error whatsoever, that would be evidence against reductionism, because there would have to be some sort of holistic thing-ness that is hiding the effects of the quark-and-electron level from higher levels.
I had a third-grade science teacher who said something very similar to me after getting a perfect score on the first test of the new school year. It did two things; it made me think she was an idiot, and it made me feel guilt. Because, see, I didn’t actually put in any study. I’d just paid basic attention in class, and then spouted it all back out.