With the Lampi DHT output section the tube's Rp (plate resistance) is the output impedance, which is generally much higher than typical and this can make a big difference in some systems depending on the preamp's input impedance. You'd also need to find the ideal bias for each tube type which isn't done afaik, so there are some factors that make tube rolling in such a circuit not necessarily applicable to everyone's situation.
Exactly. I questioned this very thing. How can you just hot swap completely different output tubes and expect them to all be working at their optimal voltages and bias currents? This alone will create significantly different sound characteristics that has nothing to do with the specific tube type. I suspect that what the GG guys are raving about with regard to the 242 tubes is nothing more than the circuit is, either on purpose or accidentally, best optimized for that tube. It is one thing to roll different makes of the same tube type but quite another to change the tube type altogether. Being "compatible with" and "optimized for" are two totally different things.
It is the same with the 6N30 and 6N6 Russian triodes. You can easily swap a 6N6 in place of a 6N30 (the other way around though might not work due to heater current draw being higher for the 6N30) and they are pretty electrically similar but not exactly the same. Often the 6N6 sounds better, smoother and a bit richer, but this might simply be that it is not optimized in a circuit designed for a 6N30...or it could be a characteristic of the tube but until you optimize the circuit for that tube you wouldn't really know.
I have been playing a lot with bias on an Aries Cerat Diana Integrated amplifier. You can adjust the bias on all three stages: input, driver and output. It makes rather significant changes to the overall sound that I can see someone perhaps preferring a higher or lower bias for varios reasons. I personally, after much fiddling, went back to the factory recommended settings as the best balanced overall (probably they are where the tubes are most linear anyway).
In that sense, I could see how a Lampizator Golden Atlantic, which is fixed in its output tube choice, could sound better than a Golden Gate with tubes that are far from their optimal set points.