On the actual accuracy of structural parameters in single-crystal diffraction studies and possible origins of their systematic errors