Tube sag doesn't really exist in these types of amplifiers, SET amplifiers are single-ended class A, meaning the DC current draw on the power supply is essentially constant, which means there is a constant voltage drop across the resistance of the power supply. By Ohm's law (V=I*R), a current going through a finite resistance will result in a known voltage drop. If the current increases, the voltage drop will increase proportionately. BUT because we are talking class A, the current draw is constant and is defined by the DC bias points of the tubes. Since the current is constant, that means the voltage drop across the rectifier and power supply is constant.
Now, if you look at a class AB push-pull amplifier, once the amplifier enters class B, the current is no longer constant! The amplifier circuit now represents a dynamic load, the current drawn from the power supply will be variable.
What does that mean in terms of the voltage drop across the power supply? If the current is changing, then the voltage must also be changing, so the bias points of the tubes will be changing. THIS is sag, the voltage of the power supply will "sag" (i.e., go down) as the amplifier draws more current, which is undesirable. This is something that needs to be accounted for in designs that have a variable DC current load.
For single-ended class A tube amps, like headphone SET amps, sag doesn't occur since the current draw is constant.
In terms of what you're hearing when you change rectifiers, it could be as I described in my other post, that the bias points are being altered when you change rectifiers. A higher voltage drop rectifier might sound "slow" compared to a lower voltage drop rectifier as the tubes are at a colder bias point (i.e., lower plate voltage, lower bias current), which means higher internal resistance, which means a slightly higher output impedance and likely higher distortion.
If you are using two rectifiers of the same type that are measurably identical in terms of their voltage drops, then I have no idea, and I will call that voodoo. Maybe material differences are causing an appreciable change in sound, maybe there is an element of placebo effect, I can't say for sure.