Here's an example of that earlier reference where the math trace product and the mean averages differ so widely that the the one can be positive where the other is negative.
Note also, in this example, the level of that spike over the shunt. It's huge.
BTW at this level we're dissipating in the region of 40 watts or thereby at the load. It is, nonetheless, not the waveform that will be disclosed at the demonstration. It will only be shown as one example of many more that need to be fully evaluated and fully researched. What we need, most urgently, is academic evaluation of all these anomalies - if such they are.