I think what Jeff is getting at but not stating is something to the effect of this: You can't go off the assumption that baselines will produce comparable results. One program may have fundamentally different processing of things which would rend even a 'baseline' comparison of things null. You are assuming that all settings at 0 on one = all settings at 0 on another.
On the other hand, I do see the thinking behind what you are doing and understand it. And, frankly, I bet your tests are more telling than we know. As I sit here I'm going through another batch of shots on the new 1.1/4.1 engine and the detail smearing (real or percieved or whatever) is killing me.
Thanks.. I understand Jeff's comment. My assumption is not that they will show comparable results with settings at zero.. in fact, I was betting on the fact that they weren't going to be comparable. Even the George Jardin video on the new sharpening built into LR 1.1 nulls out the noise reduction to see the effect of base level NR. He also recommends doing things at 2:1 so you can really see the effects.
My belief is that if the base level of one package is less detailed then the base level of another package, then the package with less detail to start with will have less detail throughout the rest of the process. I find it hard to find a flaw with that logic. My test was only to say that at their respective base levels, the C1 Pro appeared to have more detail as a starting point.
My guess is that if someone went back and did this with LR 1.0 and LR 1.1, you would see something similar, or with CS2 and ACR 3.7 and LR 1.1, etc.