Will changing gain affect frequency response?

Ghosthouse, yes, you're right that the Fletcher-Munson Effect will not cause a change in perceived tonal balance if the volume at the listener's ears remains constant.

Regarding your original question, yes it does seem conceivable that changing how the overall system gain is distributed among the various circuit stages that are in the signal path could affect perceived brightness.  Although without knowing the details of each of the designs that are involved a more specific explanation would be speculative at best. 

For example, Atmasphere has pointed out in a number of past threads that very small amounts of certain harmonic distortion components (e.g. the 5th, 7th, and 9th harmonics) can affect perceived brightness.  Changing the signal levels that are processed by the various gain stages that are involved could conceivably affect the degree to which those distortion components are generated.  Erik mentioned another example, which is that impedances might change, depending on the specific designs.  If the output impedance of a component were to change significantly, and is relatively high, its interaction with cable capacitance could affect treble response to a perceivable degree.  And it seems conceivable that changes in impedance at internal circuit points could also affect the treble, depending on the specific designs. 

Best regards,
-- Al
 

George makes a good point that sonic consequences can particularly be expected if a gain change is implemented in a circuit via a change in the value of a resistor in a feedback loop.

Regarding Dentdog’s Zesto Leto, there doesn’t seem to be any indication at the website or in the manual as to how the gain change is implemented. (Although the manual states in bold-face type, in the paragraph dealing with the internal gain switch, that "due to high internal voltages there are no user adjustable switches, only authorized representatives can make these adjustments," and also states that the "warranty is void if the product has not been used in accordance with the instructions found in this owner’s manual.")

A schematic for the McIntosh MC60 can be found on the web and shows that its gain control is simply a resistive pot (potentiometer) used as a voltage divider and connected to the input through a capacitor.

As I see it the bottom line to Dentdog’s question, and to such questions generally, is that whatever sonic effects might result from changing how the overall gain is apportioned between components figures to be highly dependent on the specific designs that are involved. So in most cases the only way to find out is to try the various alternatives in the particular system, IMO.

Does trimming the gain on the amps give me less headroom when listening at higher volumes?

No, assuming that you don’t turn the gain controls on the amps so far down that you can’t drive them to full power (when and if that is required by the music and your volume preferences).

Regarding the ZBOX, it sounds like it uses the tube in a cathode follower configuration, and has an Alps pot at its input for gain control. It is described as using "large hi-end coupling caps," so perhaps the specified output impedance of "less than 1000 ohms" is conservative, even at 20 Hz. In any event, I would think that the key to its sonic effects, in addition to possibly allowing the gains of other components in the system to be optimized, is the mention that it "works by creating a starved voltage condition in the filament of the tube that in turn causes an increase in favorable harmonics and a decrease in objectionable distortion." Although I suppose the reduction in "objectionable distortion" may not really be a reduction, but rather a masking by the "favorable harmonics" that are introduced.

John ("Ivan"), thanks for the nice words.

Regards,

-- Al