A note on the importance of including monoatomic overlap densities in the calculation of CNDO/2 charge distributions