Some remarks on the core resonance integral as used in the semi‐empirical π‐electron calculations