[sasaf5]

This is a real-valued computer, not a quantum computer. In the described algorithm the state is the real-valued angle of the polarizer. One could very well implement this algorithm using the charge on a capacitor. Also the algorithm has bugs, it can overshoot the vertical. The author does acknowledge these shortcomings in the "caveats" section. But with all those caveats, you are not building a quantum computer at home.

[jerome-jh]

The proper term would be "analog computer", but I agree with you. Really this is not even a computer.

[sasaf5]

Indeed, I should have called it "analog computation" instead.