[noobermin]

>Regardless, this goes against the author's point, because in each case 'work' needs to be done to reverse the entropy of a local system (e.g. scattered nuts) at the expense of the larger system (squirrel heat emitted into the universe)

I submit to you that you did not get the point the article is trying to make because it was exactly this. When considering the animal expending work as the system, it's entropy doesn't decrease because it isn't a closed system. You can then retort that the 2nd law concerns a larger closed system, but you can keep playing that game until the 2nd law essentially becomes a tautology, and becomes useless in understanding the system at hand.

I don't know if the author made this point explicitly (he hinted at it at the end), but one needs to actually know the details of the system under consideration, and very general laws can bring some level of context but will be limited in terms of the actual relevant or useful insight one can glean.