I see. As I'm not done with the mods I'm not sure if I will still try different cathode resistor values. Still leaning towards the 470R as it had more prominent mids than lower values.
I don't know what anode you are using, but if you are using 68k then I agree that 470R, or thereabouts, had the liveliest sound.
But for the anode resistor I remember you recommended 69K and I have 68K. What resistor brand do you have there? It seems most brands have 68K available but not 69K.
Don't get too hung up on that difference. The reason is that it will not have much, if any effect because the operating point will not move with that much of a difference in values. Also other factors will come into play, your amp will most probably have a slightly different B+, depending on the components you have put in, making the calculation different as well.
I didn't fully document my mods so I can't remember if 69k was what I arrived at in doing the calculation or if it was a specific value that had been used. I suspect that the 69k value was calculated from the load line and I then found the nearest resistor to that value to put in. Also I kept the stock 220K anode resistors in place and added another resistor in parallel to arrive at the 69k value. I think I added a 10k to get the final value of 69, or 68k or whatever was nearest.
For the anode plate load it is usual to load it many times its plate resistance which is done for the stock amp. This is good because the resulting operating point allows 6SL7 and 6SN7 tubes to be used, neither being at the optimum point though. Also the output current is low. So we tried different anode resistor values to boost the current and change the operating point to arrive at the optimum value for the operating point. I think that's why the 69/68k anode works best because it doesn't go as low as the 33k that I tried which was too low a plate load.
There's still scope to try different values to vary the operating point still though if you want different sounds as I suggested. The lower the point on the graph the more distortion the sound will have which may or not be pleasant to try giving a more tubey sound. I prefer a more transparent sound with just a touch of tubes in there.
I used TKD resistors.
Today I started wondering what would happen with more than one cathode bypass cap in parallel... for example both a Nichicon KZ and an Elna Silmic.
You can add as many bypasses as you like but we don't recommend it. The main purpose of the cathode caps is to help transient recovery which will affect bass frequencies. The fastest cap you can have will be best to do this and avoid clipping as much as possible. Apart from this it won't have that much effect on overall sound signature, that is where the coupling and WCF caps come in, they will have greatest effect on sound.
If you imagine several batteries of different types all connected then the electrical signal coming out will be mashed up because of the differing discharge rates. This won't matter ordinarily but for the music signal it needs to be as pure as possible so smearing will inevitably occur as the signal is affected by the differences in the charge/discharge cycle. The signal at one point in the chain will be trying to move on but at another part it will be waiting for another cap to finish it's part of the cycle.
Your findings on the different cathode cap values was very useful and proved that going even larger in capacitance than we did had a beneficial effect on bass responsiveness! I suspect that going even larger than that would have negative returns possibly.
The whole point of much of the mods is to get as stable PSU as possible. The large fast capacitance MKP decoupling caps you put in will help as does the fact that the amp is balanced in the first place. Also we uprated the PSU caps to be as beefy as possible to provide a good reservoir of power on tap. Apart from that we can't do much with the PS without completely rebuilding it which would involve adding on an extra stage to the amp as it is now.
Where we did go for a bypass to the cathode cap was because we swopped the stock caps for higher capacitance, musical grade caps, which meant that we needed to add some very small bypasses to complement the lytic cathode cap techno. So we added some small fast film caps to restore the high frequencies that were lost.
So I would suggest if you want different sounds to try different coupling and WCF caps. Due to their position in the push/pull cycle the WCF caps we feel need to be the fastest and responsive to the high frequencies as you can go for, I have Mundorf SIO which I feel complement the more natural sounding Jupiters in the coupling position.
The switch you mentioned before would be easiest to implement for the different anode/cathodes by adding a supplementary circuit with other anode/cathode resistors in parallel. With a switch for different coupling caps that would be more complicated to implement though.