You can experiment with most of the small power pentodes. I liked 6g6g. The setup used here are well within the margins of their limits anode wise so no worries there, just find your proper cathode resistor for the tube and listen away with different anode voltages. Look out for the heaters to so you dont fry them. Your biggest enemy will be microphonics.
Wima FKP are very good film & foil capacitors. In any case when deciding a coupling cap, beyond calculating a uF value for enough low pass given an amp's input impedance, we must also remember that the V rating of the cap should allow not only for the DC it blocks but for DCV+AC it sees. 1.414*Vrms +DCV <Vr. That is why we see a 100V marked capacitor rated for 63V pure AC in some analytical datasheets for instance. Because 63VAC*1.414 = 89V and then +10% allowance for safety and longevity. That's for line and low enough frequencies mainly. Especially in pulse HF use the % allowance is much more as the V rating goes up. There the high energy changing steep and fast can be much more damaging to dielectrics.
Merry Christmas !
Would by-passing the original cap with a smaller cap but with much higher voltage rating help the situation ??
Bypassing can make a combination speedier but it also mixes in more than one part's different self resonance and introduces different ESL due to more legs. For audio frequencies coupling its not technically critical, more a matter of subjective trial, unless someone has adequate electronics lab equipment and knows how to measure and evaluate such things.