Convert mA to A before P = V x I. For each constant-power interval use E = P x t, then add energies and durations for the successive run. Average power is total energy divided by total time, not the sum of powers or an unweighted average when durations differ. The declared ideal readings do not establish temperatures or safe equipment. No real circuit or waiting is needed.
Worked example
A separate example has 1 W for 3 seconds, then 2 W for 1 second. It transfers 3 J + 2 J = 5 J in 4 seconds, so average power is 5 / 4 = 1.25 W. Neither the sum 3 W nor the unweighted mean 1.5 W describes that full run.
Notebook: one run with two successive intervals
Interval A lasts the first 4 seconds: a resistor has 4 V across it and 75 mA through it. Interval B lasts the next 2 seconds: a different resistor setting has 4 V across it and 150 mA through it. Values stay constant within each interval. There is no additional time or energy between them. These are paper records of an ideal model, not a physical experiment.