It’s not too difficult to show on-paper, the limited Cp of a Savonius design. Since the working surfaces are traveling downwind at half the windspeed, you can expect about 1/4 of the power one would hope for, from this downwind-traveling, power-producing side. Then subtract the power used to push the upwind-traveling side into the wind, and you are probably down to between about 1/6th and 1/8th of the power available to the swept area. That agrees with the measured Cp’s we typically see of around 0.1 or less for Savonius.
Still, any power is better than no power, and if it works and survives, there could be some use for such a configuration. And the two-blade overlap version supposedly reduces drag on the upwind-traveling side, although it also reduces the overall swept area, travel distance, and leverage from the working surfaces. Forgive my bad memory, but I’m not quite clear on whether a Savonius generates much elevating lift, in the manner of a magnus/flettner spinning cylinder, or a Sharp rotor. That’s why I asked if that photo was “line-laundry” (hanging from a kite, located above) or a free-spinning, self-elevating kite in itself. I’m guessing it was line-laundry, dependent on an elevating kite to support it. You may find you also need a kite above it, to keep it elevated.
I would point out at this point, since this AWE hype-cycle began 15 years ago, kite-flyers have naively thought spinning line-laundry would be a slam-dunk easy way to generate electricity using off-the-shelf kite-festival components. So far it has never turned out to be true. Any “real wind person” could have told them that. And I did. But that didn’t stop them from trying.
As I remember, the Magenn MARS failure made very little power - a typical “Professor Crackpot” design. After appearing on so many magazine covers and websites, including NASA, it quickly turned from the poster-child for AWE, into a complete joke (for the few people who paid attention beyond the “press-release breakthrough” stage of hype). That didn’t stop people from continuing to cite it as a great advance. I think I recently posted a link to a new video announcing this old, disproven concept as a new breakthrough! (More clickbait.) Once Professor Crackpot gets started, it’s difficult to extinguish his bad ideas from being repeatedly promoted!
The size of a large bus or more, Magenn/MARS supposedly hit about 1 kW or so? Using bicycle wheels and fan belts or something? It could have probably won a competition for the least effective wind turbine ever built, using the most material, at the highest cost, to produce the least power. No wait - that would be four (4) competitions it could have won at once.
Anyway, I will also point out, as admirable as your quick-and-dirty, “get 'er done”, low-cost experiments are, you’d probably at some point need to build a high-quality version to get any meaningful numbers beyond the typical “lighting an LED” stage of verification. 