I've noticed that when gas-filled photocells were used in valve equipment, they were nearly always supplied with a low-impedance source of +ve voltage to the anode and the signal was taken off a resistor in the negative return. There is a blocking capacitor between the photocell cathode and the grid of the valve, so the standing current and DC conditions don't appear to be relevant.
This means the photocell has to be connected by a 2-core screened cable, which was an expensive luxury in those days. It also has to be thoroughly screened to prevent hum, whereas the cathode half-cylinder would partly screen the anode and reduce the amount of extra screening needed; so what was the advantage of taking the signal from the cathode instead of the anode"?