How does it reveal this? All results are somewhat Dac dependent. Over at AudioScience Review Amir has had fun pointing out that various foo streaming devices have worse results than his laptop, but
a. the differences are inaudible
b. they depend on the method of powering the foo streamer bad psu tend to be the cause- but note this does not mean cheap vs expensive or linear vs switchmode, although.
c. they reduce or disappear with a competent dac.
It's clearly audible. I don't even use super-dupa PSUs, just a pair of ifi ipower 5v's and I can still hear the difference.
Measurements for the Allo USBridge are public on their website - its competent alright. As for Amir, he sometimes falls into the same trap as archimago - not all that competent, has an agenda, but make of it what you will.
Amir doesn't come from an actual eletrical engineering background either, and it shows at times. But I wont entirely shoot him down - there's some merit in what he does for sure.