I don't recall anything having been said during the previous similar incidents that would link re-appearance to anything other than local circumstances on our own machines.
Be that as it may, I doubt this 'breakthrough' of the favicon was (just) a local phenomenon. Looks like something upstream (cache?) contributed, and perhaps we'll find out more. For now things work, at last.
