[hyporthogon]

(1) 'Exchange between strangers is a unique feature of us modern hominids' isn't really falsifiable without begging the question or establishing a hard boundary between kin and non-kin, which afaik is currently a function of the problem under consideration (i.e.'kin' might mean something different to geneticist vs. anthropologist vs. linguist vs. classical economist and even something different for a different problem in each of those disciplines). But 'this is what makes us human' is far from essential to Hayek..and probably basically inimical to Hayek, as some other commenters have pointed out.

(2) The 'reduction of collective intelligence to the price mechanism' objection (which is often a specific case of generic objections to dimensional reduction, including e.g. perceptron thresholds) is addressed throughout Minsky's Society of Mind[0], esp. the 'frames' concept.

(3) The article doesn't mention the benefits of localized knowledge (as opposed to the 'practical reality' of localized knowledge, which may be lamentable and/or fixable), which iirc Hayek does get into (or maybe some other Hayekians? more modern systems-oriented folks? can't think of a source at the moment). If some knowledge (for example, within a community of practice) weren't pretty strongly localized, then every knowable would be in one truly global variable space, and abstraction would be incredibly computation-intensive, and knowledge growth would be horribly O(n!), and all of thought would work like JavaScript (ZING). This is a stronger kind of localization than technical specialization (which just maps onto SOLID class design rather than variable space).[1]

[0] Beautiful html edition: http://aurellem.org/society-of-mind/ [1] Broad philosophical musing on this sort of thing: https://mitpress.mit.edu/books/bubbles