I do use speaker wire. pure copper for low resistance, and there are many thicknesses. I used 0.75mm2 before (and it works). but now use 1,5mm2 12,5 meter, total 25 meter. for increased capcitance and inductance, with decreased resistance.
I'm not familiar with turbine effect.
the impulse gives a longitudinal push (like a tsunami)
I'm still looking into the impulse speed, I read abou the 1uS in " cold war technology", made a video series about that book.
peak to peak? there is only one peak, as it is unipolar.
I measure (in my setup I just tuned with 330pF) a 750nS period, where the impulse goes from 0volts to -650V to 0 volts
The turbine effect has to do with some of the replications of Tesla's boundary layer turbine, I have seen a few videos of people displaying an odd sound and massive acceleration after a point (usually in the tens of thousands of rpms, lower for broader turbine blades) so they can do something very similar to your work where a fairly small input brings the turbine up to a speed where there is a sudden disproportionate voltage/pressure rise.
I think the transition point is when it goes from merely being a spinning fluid (whether air or aether) to this vortex effect.
I bring this up, not to distract from your work, but to present a slower speed example of the same phenomena in a fluid of differing density.
https://youtu.be/XVW44ZZa62Q?t=67
I would compare the portion before the jump to pulses too slow to achieve the effect you are creating, they still accelerate the system, but nothing too exotic.
At about 1:42 in the video the air powered Tesla turbine crosses the 50,000 rpm mark and suddenly accelerates by more than ten thousand rpms in a very short span.
I think this is the same fluid dynamic effect your hairpin circuit is causing, but you are doing this with a much more tenuous fluid than air so the cycle times have to be all the faster.
This suggests to me that the turn-on delay and the rise time of the MOSFET are the most important timing factors of a device using these principles. It is my hypothesis that a slower overall frequency might be fine so long as the rise time is still short enough and the pulses are kept sharp enough. I think the duration required is inversely proportional to the mass, where a larger mass can have a longer pulse rise time and still get the desirable effect.
I have found your posts very helpful, thank you.