At an ISP of <200 I'm surprised they are bothering with the microwave tech. It would be more efficient to separate the water into H2/O2 and then burn it in the traditional manner.
It turns out that electrolysis in zero-g is tricky, the bubbles tend to stick to the electrodes. I believe there are some nanomaterials in development now to prevent that, but it makes it more complex than you would initially think.
Then don't do it in zero g. Spin the chamber and siphon off the combined gasses from the center. It might not even need 'moving' parts. The contents of the tank could be stirred with magnets.