Kontsevich and colleagues' logic seems to be the following: posing the unique criterion constraint (their eq. 2) and given the equality of the zFAs (or, equivalently, of p(FA)-s; their eq. 3), the ...