Remembering it is a pre amplifier/ passive attenuator that is being measured, not a amplifier. End users would likely never use full volume at all with levels of 4V, not even at a Black Sabbath reunion party.
It can be seen full volume tests with 4v in and 4v out, mainly just tests the devices wiring that it has in place input to output, and indeed you would expect, as seen in the graphs, excellent distortion figures are the result, being a wire and a RCA plug or two and a attenuator full on and fancy gear measuring.
What is needed then in the future , is measurements indicative of actual use as a volume control, over many levels as this is how the product,if its an attenuator is going to be used by consumers.
IMO, measurements for evaluation of passive attenuators needs to occur much lower, and I would suggest 50mv to 500mv is more indicative of actual output levels occurring in consumer source equipment when playing regular music. The times I have measured AC music signals from the output of a CD player I have not seen any at all above 290mv
If I am wrong with this, does anyone have a voltage graph, to show min to maximum voltage ranges, of a standard CD player playing lets say Getz /Gilberto ? ....er how did they get such a good recording - particularly the vocals.
The exception though might be Oldfield's Amarok that momentarily fully tests a CD players output ability, and certainly higher levels can occur momentarily perhaps to 1.5v over a second or two's duration with such pieces of music, but these moments on CD's are very rare indeed, inferring level testing should IMO be much lower.