Some remarks on a numerical representation theorem by Newman and Read