In a pure 10-ohm resistor with a constant voltage of 10 volts flowing through it, the current will be 1 ampere. As in your example, the DTC is 25%, so the average power, which when using a pure ohmic resistor, will also be equal to the instantaneous power, therefore 2.5 watts. This is verified by your formulas, and it is normal in this case of power with a pulsed circuit using a pure resistor for the power to be constant.
Since we are talking about the ZPM and we are using components that store and release energy (coils), in this case, the instantaneous power will be different from the average power because the current through the inductor increases from 0 amperes. The initial instantaneous power will therefore be from 0 watts to a maximum of 4.33 watts, as in the following example.
Experiment already performed on the ZPM with this figure: Yellow PWM pulse , Green Current probe

On the oscilloscope, we can see that the current increases linearly with a starting voltage of 0 amperes.

The PS tells us that 4.33 watts comes out of it making an average of 4.33 Watts, yet at the beginning we have 0 amps and therefore 0 watts of power, it interprets a maximum power as an average power. The PS therefore makes an average power of 24 x .18 = 4.33 watts it does not know that this power is transmitted only 38.2% of the time starting at 0 watts and therefore considers the average at 4.33 watts random and not precise reading, as would a DDM connected to the input.
The power is not constant and we cannot trust the display of the DDM of the PS, they do not have the same formulas to use for the calculation of power and a DDM of PS cannot do that in addition to the frequency used. With an oscilloscope the average power displayed will be different, because it can do Pave = Vrms x Irms x DTC
The ZPM is beyond unity for sure.
Jagau