The original topic had to do with tube amps sounding more like transistors.
The problem is that there is a model for how speakers are to be driven by amplifiers- this model I call the Voltage Paradigm. In this model, the amplifier behaves like a voltage source, which is to say that for any given impedance it is to drive, the amp will make the same voltage. We are all familiar with the ideal here- as you cut the impedance in half, you double the power of the amp. Such an amp is a perfect voltage source.
There is only one way to do this with tube amplifiers: add negative feedback, probably 20db or more. The problem here is that all amplifiers have a propagation delay (the amount of time it takes for a signal to propagate from the input to output of the amp). What this means is that the feedback signal, usually taken at the output of the amp, will be slightly delayed, IOW will arrive slightly late back at the input of the amp, to really do its job correctly. You can see that this problem gets worse at higher frequencies due to the shorter duration of high frequency signals.
(BTW the amp will not double power as impedance is halved, it will cut its power as impedance is doubled. So whatever power it is capable of into a lower impedance like 4 ohms, will be the maximum power that the amp makes.)
Because of this error, the addition of negative feedback has some adverse effects. One of them is increased odd-ordered harmonic distortion, something that is usually associated with transistor amps. The ear finds such distortion to be an increase in brightness and harshness, as the ear/brain system interprets distortion as tonality. In this case, the ear/brain system also uses the higher ordered harmonics as loudness cues. IOW uses these harmonics to sort out how loud a sound actually is. When there are increased amounts of higher ordered harmonics, we perceive the result as harsher and brighter.
Hence, tube amps with lots of feedback will sound solid state.
Now there is another model used to drive loudspeakers that has been around for a very long time. I call it the Power Paradigm. In this model, its expected that the amplifier will make constant power rather than constant voltage. Obviously no amplifier can do this (but neither can any amplifier behave as a perfect voltage source either) but some can come quite close! This requires different design rules in the loudspeaker, as the crossovers will not work the same way and the box resonances are dealt with in a different fashion.
Now what most audiophiles do not realize and what the industry does not like to talk about is the fact that these two models have been in competition with each other since the late 1950s when Macintosh and EV first proposed the voltage-drive rules. Most tube amplifiers that do not use feedback are on the Power Paradigm. Loudspeakers that work with them (most horns, ESLs and a good number of box speakers including the venerable Acoustic Research AR-1, the world's first acoustic suspension loudspeaker) are in the Power Paradigm as well. Part of the equipment matching conversation we see so often in audio has to do with the fact that you can't mix equipment from differing models all that well- you usually wind up with a tonal coloration; no bass, too bright, stuff like that.
More information:
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php