Good question. Peter's SS circuit is doing it, but not in an orderly fashion. Any time-dependent electrical field that has a high enough amplitude and is changing. Happens in the solar system all the time.
My honest opinion:
The coil is doing it to the core conductor. There is almost no indication of such induction if the spike isn't sharp and short. It is a simple matter of orientation.
In normal induction both wires are oriented the same way to the fields of the other. Yea, you can twist the iron core around and make it look different and still work but the the magnetic field of the core is in the middle doing the twisting.
In-short... normal induction works when the primary and secondary are parallel to one another.
Peter's is different. It is more like a knife cutting through the wire while perpendicular to that wire. As long as that is the only action going on he will only see very short spikes. These spike are from the mag field of the CC turn cutting through the cross-section of the core wire. This will generate almost no usable energy and the spikes will be positive and negative (field passing through the core and then returning).
Obviously, there is more to his results.... the explosions and following bursts.
I believe the coil wrapped the length of the core is to provide the magnetic bias to help make the A field unidirectional. The coil that should be pulsed to perp induce into the core should be elongated or elliptical. That core cannot be equidistant from all points of the inner surface of the pulsed coil because dead center is the nulling point for those pulses.
Each core should be wrapped as Peter's is. But, there should be two or more such coils. Each at extreme ends of the widest inner part of a larger pulsed elliptical overwrap.
I have not confirmed if those loops should be individual loops or two loops folded from one.
This is why I cannot believe the SS matters except to start the process and keep it under control. The sharper and shorter time for the pulses the better. Once the energy becomes unidirectional the gaps would be filled by new pulses. However, this is probably near the level of the latest TPUs seen.
I'll be very happy if I can make one work that required magnets and I don't mean for tripping switches.
I very much doubt any high energy electronics is required.
In any case, it is very good different directions are traveled. Someone is sure to find the right path.
Peter has the high energy anomalies. I have a steady but very low DC output with rectifier noise on top. I could probably light a LED if I used a joule thief but it would never power the multivibrator generating red noise for the perp induction. A dead end for me. It is like a shadow of a virtual N-machine that will never do better.
So back to the board etching and rectifier garbage output experiments on my bench. My priority is to duplicate Peter's results so I can run tests with equipment I use for work. It is very slow working through an 8" magnifying glass - my problem not the circuit's.