I just noticed that some equal sign should be substitute by semi-equal sign ≈
Though the probability is not exactly equal to what we have after Stirling expansion, the convergent/ divergent behavior is what matters, and is correctly described by one over square root of n for large n.
