It's funny to see this show up now. Optical-domain switching was a hot area of research in the early 2000s (e.g. [1]); it largely petered out as it became clear that digital-domain switching was more practical (because it didn't rely on exotic technology, and provided time for hardware to perform routing operations).

[1]: https://www.laserfocusworld.com/optics/article/16556781/many...

Is optical digital or analog?

Either/both. A nice feature of lightpath networks is that they are somewhat independent of the signal encoding.

Optical computation can also be digital or analog.

In this case, the control signal appears to be digital while the switched signal can be anything.

Can [such] optical signals be used for photonic quantum computing; is it only unitary transformations on the photonic signal?

Google (and I'm sure all others) use Optical Switches for years

https://cloud.google.com/blog/topics/systems/the-evolution-o...

That's an electrically controlled optical switch, whereas this is an optically controlled optical switch.

Still very cool though, I didn't know the former existed.

I'm not sure I fully understand how one can convert different twist directions into an clear output. Does CCW vs CW change where the end of the affected beam will end?

My first guess is that the change on the electronic band allows downstream electrical switching of a prism/mirror or other photonic device. So you have a special light signal controlling an electric signal to switch the “regular” light signal. Not sure about the applications for the strong EM field though.

Have any 2D thin film materials been commercialized yet?