On the role of fractional calculus in electromagnetic theory