I read the entire build and I agree with you. Well, what I have been saying for 20 years and what your results show are basically the same. I think you took my comment out of context, I'm not saying the 1 3/4" are a restriction, I was only saying that if Cuda was going to change headers down the track, why fit the 1 3/4" in the first place when there is very little difference in them. Your results from PVF1 as far as the header change went doesn't really reflect matching headers to an application. It looks like the heads were specifically flowed and ported to suit 1 3/4's. Just by changing the headers, it would have changed the flow characteristics of the heads. Same the other way, if the heads were built to suit 1 7/8" headers would you have seen a similar result stepping down to 1 3/4"?
Another thing I have been saying, and most dynos show it, is that headers are only optimal in a given RPM range. Can headers "over scavenge"? I don't think so, it's more likely the primary size and length isn't matched to the RPM range that the power is dropping away. Here is a dyno of an LS1 that I custom made a set of 1 3/4" proper tuned 4-1's to suit the application IE: burnout car. The specs of the pipes were matched for 4000-5500 rpm. The torque curve shows that.
View attachment 196203