It seems like the biggest arguments against UUIDs in that post don't apply here, and the other arguments are explicitly acknowledged in the "caveats" section of

It seems like the biggest arguments against UUIDs in that post don't apply here, and the other arguments are explicitly acknowledged in the "caveats" section of OP. Also half the things in that post seem like they have nothing to do specifically with UUIDs at all (scaling, foreign keys, exposing primary keys)