On the fidelity of the core mass function