Even there. I’d suggest that if a difference was detected, the most probable hypothesis would be that the difference was insufficiently hidden from the test subjects. The best test would also remove the wildly variable and unrepeatable instrument of human hearing from the experiment.
And really, it’s pointless. What is it people think happens to the current fed into a digital computer with an external power supply? What happens is that it’s fed directly into a PMIC, a power management integrated circuit, which turns it into a set of switched-mode voltages to power the various components of the computer. Yes, and to power the USB hubs, if the computer has any. So it’s all switched-mode power anyway in the places where the misinformed think that it’s important.
The Raspberry Pi, for instance, uses this chip. If you look at the data sheet there, you’ll see “1MHz to 2.1MHz switching frequency”. You’ll also see that it contains four “buck regulators”, each of which is a switching regulator, to produce the various voltages the computer uses internally.
So all this guff about putting linear power supplies on digital computers (or any Raspberry Pi or equivalent audio devices) does not accomplish the linear effects the clueless are hoping for. Computers are not built with vacuum tubes anymore, and zeppelins no longer moor atop the Empire State Building.