Switching out the cathodes of 2 power valves for half power - switch danger?

[Cirrus]

How worried should I be about voltage on the switch of a half power switch set up like Matchless did their half power switching - a DPDT on-on switch that both switches in appropriate extra cathode resistance to bias for 2 valve operation, and lifts the cathodes on the outer two el84s to defeat them?

I was just thinking that with the cathodes lifted, the whole of each unused el84, and the wire from the cathode to the switch terminal, will be floating at the B+ voltage, right? So how worried should I be about that switch, with 335v across it?

It's a flat recessed slider switch that needs a screwdriver to move it

[sluckey]

I would have zero worries.

[pdf64]

With their g1 still referenced to 0V common, the open circuit cathodes would probably drift up the a VDC required to put them into total cut off.
Much lower than the HT, probably more like 40V with EL84.
To help ensure that, maybe fit 100k pull down resistors to them (either across the switch contacts to the cathode resistor, or to 0V common). 
That would probably be good practice actually, as there’s the h-k voltage limit to consider.

[PRR]

If G1 is still held to zero V, the space "under G1" (the cathode) will be below Va/Mu. In this case "Va" is really the screen grid and Mu should be taken as Mu(g2) which for EL84 is about 18.

So 350V/18 is 20V, maybe 30V or 40V since Mu falls-off as current vanishes.

Not 350V.

An observation: that type switch is usually used ONCE, when an appliance is sold in a different market than what it was made for. (Not even then, usually.) I think it is the same guts as "standard slide switch", but those have given me plenty trouble. Be prepared to replace.