A good article, Matt. Stafford Beers was, to the best of my knowledge, the first person really to elucidate the idea of emergent properties (he called this, somewhat confusingly for today’s lingo, “cybernetics”). Since then people like Peter Checkland at the University of Lancaster have elaborated the idea with regards to intentionally designed systems, but the concept is valid across all domains, as you correctly point out.
I think the one area in your article you’ve not quite achieved clarity about is the quest of physicists for a unified theory. What’s being sought isn’t a reductive “answer” that’s at the Planck scale but rather a theory that stretches from the very smallest unit to the very largest. By way of example (from another discipline) evolutionary biologists have for a while been puzzled by the fact that evolution ought to move more slowly than it appears to do in reality. Lately, biologists working with physicists and mathematicians have realized that there are “phase spaces” rather than individual molecules; thus any molecule in a given phase space may act as the locus for change. In other words, instead of evolution having to rely on the random mutation of a particular gene (which is read to produce one or more protein molecules) it can work with the mutation of any one of hundreds or sometimes thousands of genes, because the basic configuration of the proteins they code for is within the same phase space.
This is an example of a general rule applying from microscopic (protein) to macroscopic (an entire species). It may well be that such phase spaces pertain to all phenomenon; this is, in essence, the underlying idea behind string theory although string theory itself is unlikely for various reasons to prove out. Thus the concept itself would span all levels of emergent property, simply giving different “results” depending on the precise level of abstraction required.