We need a blind test to sample a full feedback design vs. the no feedback design.
Maybe use a decent budget mm cart like an AT Vm95en or At vm95ml, and compare a few short samples using the Pyxie and a comparably priced discrete design such as the hagerman bugle or the Darlington labs mm5.
It may help to point out the sonic differences between the two design philosophies - assuming the differences can be discerned.