All models are wrong, some models are useful.

I could start by dissecting the article and detail the sloppyness in the reasoning, the strawman of ignoring emergence, the value and inevitability of referential semantics etc, but ultimately the deeper question remains:

How is the proposed model supposed to be useful?