If the generator is sufficiently large and needs a strong wind to work, there is no problem with this.
There is not just the 5.5 W from driving the prop, but also 11 W of wind power. 16.5 W in and 11 W out sounds feasable, though maybe demanding on the prop quality. With less extra energy the demands on the prop get lower, so that even the relatively simply toy model on the treadmill worked.
Using some energy input to later get out more power is common: The combustion engines usually work this way, by first compressing air to than get back more energy from the expansion of the hotter gas.
The calculation does not include an extra form of energy storrage. The prop driven from the wheel mechanism works without it.
For the experiment we have not found any good hint that energy storrage is relevant:
There is no extra motor - so no battery to use.
Kinetic energy can only be used when going slower.
The slightly more tricky to analyse pressure differential idea also does not work, with two strong aguments found against it:
a) the fast decay
b) similar to the kinetic energy the pressure amplitude is linked to the speed.
Eleastic energy (e.g. in a chain) would not last very long and there is no visible change in the vehicle configuration (e.g. change in the length) while going at the speed of the wind.
The rubber band example showed elastic energy storrage, but with the energy storrage still accumulationg while at the speed of the simulated wind.
The only way you get that 11W on the output is if you add another power supply in parallel or series with the first one.
Yes, that's the principle. The propeller and the wind are 'in series', so to speak. If you like bad analogies, you can compare the various 'speed' of the players with voltage. If the wind and the vehicle are at the same speed you can compare that to a power source and a sink--say a charger and battery--that are both at the same voltage so that no power flows. But if you add an isolated DC-DC converter across the source and then put its output in series between the source and sink, now power will flow.
So the wheel-to-propeller link can be considered as a sort of boost converter that allows power to go from an original source that is the same or even lower voltage than the sink.
I like that analogy, but the DCDC converter would be more like a step down one to get more current, though at a lower voltage.
Going at the speed of wind would be equivalent to charing one 12 V battery with another 12 V battery. It does not work directly, but it works if an extra DCDC converter is used to add some (e.g. 2 V) to the voltage.