I have selenium stack in my TransOceanic radios. I'll power one of them up and measure across it. As I recall, the voltage drop was not huge - like a few volts, maybe 10. It would be similar to a series string of germanium or silicon diodes, each having 0.3v to 0.7v across it, each adding to the total voltage drop.

You could mimic this behavior with a series string of more modern semiconductor rectifiers or a diode and resistor. If the circuit is low voltage, like a filament circuit, I might be inclined to do that. But if it's the B+ supply, I think that the DC voltage desired is probably high enough that a few volts loss isn't required for proper operation. In that case, I think you probably could replace it with a single diode having less than a volt drop across it.