Note here that the deviation from flat is almost non-existent above 1khz. This was my goal. The SPL deviation is zero from the first graph.

You may be asking why I would do this at all, given the result seen on this page. The answer is as an exercise in the audibility of impedance variations with good solid state amps. I was sure that it would exist with tube amps.