Sorry, I mispoke. I was referring to the video signal itself, and didn't mean to "assert" anything.
"The same video signal decoded 3 different ways will wield 3 different results, in this case, superior sharpness and chrominance when delivered via component outputs."
Let me restate my question, then. Using S-video from LD players usually doesn't give better results than that from the composite output, generally because the S-video comb filter in current display devices outperforms the older comb filters in the LD players. I'm curious to know how the component-mod in your player compares to the eventual component signal that would eventually be derived from the composite connection. Does it somehow utilize the component-video circuitry already in the 91, or is it a separate device altogether?