This is too simplistic a view, and I was thinking specifically just related to the basic output stage which does typically behave much like a voltage source, and is usually configured as a voltage follower, and with a light load (lighter than a speaker), behaves as a voltage source, and with load, as a voltage source with an element of constant and variable impedance.@roberttdid I've yet to see a tube output section where on its own without feedback, is able to behave as a voltage source. If you can point me to one I would be very interested. You might want to take a look at this image:https://www.radiomuseum.org/r/fisher_80_az80az.html
This is a Fisher 80-AZ, typical of a number of amps from this period of the mid-late 1950s, prior to when the voltage rules were adapted by the audio industry. It is equipped with a Damping Control, which is a variable voltage and current feedback system.
Note that at 12:Noon the control is marked 'Constant power'. At the extremes the control is allowing the amp to be a voltage source or a current source, as the control operates both forms of feedback. When the two feedbacks means are balanced against each other, that is about the same as zero feedback, hence 'constant power'. Now if you spend time with zero feedback tube amplifiers, and happen to have measured their power response with respect to load, you find that above a certain impedance (depending on the overall output impedance of the circuit) the power decreases quite slowly as impedance is increased, in fact doubling the impedance sees only a small percentage loss of power. Its not perfect of course, but 'constant power' is really not a stretch; a zero feedback tube amp will do pretty well with this as long as the load impedance is high enough. No amplifier is perfect of course and this includes all amps that behave as voltage sources as well.
So my description as not too simplistic. It was simply correct.
Their response w.r.t. voltage, is fairly flat from mids-highs, with usually a bit of a dip at high frequencies. An amplifier that doubles in power as the impedance is squared will keep the most consistent anechoic output.Huh?? What kind of amp doubles power as impedance is squared? Even a constant current amp only doubles power as impedance is doubled. At any rate this statement is entirely false, as ESLs don't do that. Here's a rather famous ESL impedance curve, the Quad ESL57:
http://www.quadesl.com/quad_main.shtml
You can see that while it does flatten a bit in part of the midrange, its on the decrease all the way from the peak in the bass. We have a lot of customers with Quads and Sound Labs (80% of our MA-2s built over the last 30 years are running on Sound Labs); these speakers don't seem to behave around voltage rules nor should they, as their impedance curve is not that of a driver in a box with its attendant resonance. This is of course not the only example of a modern high end audio loudspeaker that doesn't use the voltage rules; keep in mind that most SETs are zero feedback and so tend to behave more as power sources than voltage sources, and yet there are speakers on which they do quite well as the designer of the loudspeaker intended that it be that way.