If that's an 8 ohm speaker then it's current will likely dominate the current through the inductor, assuming the reactive components have a much higher impedance at audio frequencies. The high frequency current components from the switching frequency should be relatively small and would contribute little to the inductor power dissipation.
So why is this wrong : ���� =��/(2∗��∗��
..?? and PD = IL*IL*DC resistance???