"* they did pretty much all of the things you would like to see (such as reversing the direction and making sure the thrust reverses).
* they seem to have done a thoughtful and careful job, including testing in vacuum."
Read the article carefully.
They did not actually test in vaccum. They tested at atmospheric pressure, because they did not have suitable vacuum rated amplifiers.
Spending half a page explaining how the vacuum system worked, only to have a throwaway line later in the paper (search on electrolytic) that they diddn't
actually use it is at best shoddy writing.
To quote from an earlier post I made on this.
The net torque is zero - yes.
The problem is that because the 'vacuum' chamber wasn't part of the measured system, you can exert torques against it without issue. Convection can do this and distort the measurement.
A major reason why this can't be true - or if it is it's bigger than any Nobel Prize-winners discovery in history, and maybe all of them:
The reported thrust in the NASA paper is 0.4N/kW.
Power = force * velocity.
If you put this on a railway car going at 10m/s, then you get 0.4W*10m/s = 4W out for 1000W in.
If the car is going at 100m/s, it's 40W.
At 3000m/s, 1200W.
You take 1000W of this to run the engine, and you now have 200W of free energy.
This can be arbitrarily scaled up.
If it works, it is not only a space drive, it's a perpetual motion machine that needs no fuel and emits energy.