Here I repeat yesterday's demonstration, but I include the display of the energy integral computed by the LeCroy 104Xi scope's math function. The computation is Integral (voltage x current) dt, and so shows the energy passing the current measuring point and its direction of flow. The current is being monitored as the voltage drop across the 1-Ohm resistor, and the voltage is monitored by the Differential Voltage probe.
Moving the Current Monitoring reference lead by about 3 inches along a bit of wire causes a radical change in the high-frequency components of the JT pulse. This kind of change in peak values will tend to confuse digital multimeters (DMMs), especially the cheap... er, inexpensive ones. But the oscilloscope isn't fooled: the energy integral is about the same in both cases. This would NOT be the case if the peak values were used in a manual multiplication for power, then integrated on a spreadsheet using wider time slices. That is, the peak value/meter/manual method is a lot more error-prone than using an oscilloscope with good built-in math capability.