To be honest, there are a lot of inaccurate statements in this thread. I am guessing that only a few people here know design and have experience in building an audio component.
Are there really inaccurate/false statement in this thread, or is it just perspectives and experiences that don’t match with yours? If it’s truly a false statement, it can be decisively refuted with math & physics.
One of the common perspectives I disagree with is (paraphrasing) - "doubling power is only 3dB and that’s not much anyways". 3dB is a lot! You can TOTALLY hear a huge difference in 3dB. Find your "preferred" listening volume, then adjust up or down by 3dB - it sucks! 1dB was loosely determined as the "audibility threshold" a long time ago, but the subtext there is *for casual listeners*. It’s more meaningful for a seasoned audiophile. When dealing with say L/R channel imbalances, you can certainly notice down to a third of a decibel, at least.
Another perspective I don’t share is that paralleling multiple outputs (tubes, transistors) imposes some kind of sonic penalty. Definitely NOT in my experience. If anything, the aberrant characteristics of one particular device can get averaged/smoothed out by the others (that’s good). The choice of device itself matters a lot (of course), but if the PSU and OPT and driver circuitry can handle it then the more the merrier I say. Bridging is a different story, as is stacking extra gain stages - both of these can certainly impose performance penalties in the pursuit of higher power and higher volumes.
I definitely get folks who prefer the sonic character of say 6L6 or EL34 (I love both of those tubes) over the more austere KT88/KT120, but I bet a properly done high power amp with lots of paralleled (say) 6L6GC would be pretty amazing - you just don’t see those on the market. Manley does have their NeoClassic with lots of EL34, but I haven’t heard it yet - looks interesting.