Albert: You say Krell confirmed your finding that "overhead halogen lights, shining through the Plexiglas lid of their MD10 CD transport, affected the bit stream to the D-to-A converter." This implies that they actually found a change in the measured datastream, and also that they controlled for any effect that such lighting fixtures may have on the powerline (presumably by measuring the data with the lights on, versus with the lights still on, but blocked from shining into the player).
One of our music night sessions I accidentally improved my Krells performance by resting a record album cover on it's top.
The CD was still playing and I was putting the LP on the turntable and we all heard the improvement. At first we all thought the tiny additional weight of the cardboard was somehow damping the chassis.
After repeated tests we determined the same effect could be had by holding any light blocking device above the player even when we were not touching the player or turning the lights off.
I told Krell about my experience and at first they did not believe me, later they acknowledged there was an effect and they treated the interior of the player with light absorbing material and added a heavily smoke colored tint to the Plexiglas.
Oddly enough, a guy my audio group has a very late model Krell player and while visiting him a month ago I ask for a magazine. He brought it in the room, we continued to listen and then I laid it on top, blocking his halogen room lighting.
There was an improvement and he was totally surprised. Not nearly as big as with my old Krell but audible enough that he cut a fancy piece of cardboard to fit the player and now keeps it handy when listening.
People love to make fun of tweaks but often it's just an improvement by accident that solves a problem nobody even knows exists.