I don't know of any such study. Here's an idea: why don't you do it!
I did do a few experiments last summer on saddle material as part of the runnup to my 'string' talk at the ASIA meeting. I made saddles of bone and HDPE for my 'test mule' classical and a Dread that one of my students loaned me. We swapped them out and tried some informal listening tests, and I also tried to get some measurements of how the output of the guitars differed.
What was interesting was that, although you could hear differences, I couldn't measure anything consistently different. To me the audible differences seemed pretty small, considering how great difference there was in the materials.
My thought was that the softer plastic would 'eat' some of the highs, and maybe also give less prominance to the octave-doubled signal from the tension change in the strings by bending as it was pulled on. I used a 'wire pull' technique to pluck strings in the same way in the same place with both saddles, and looked carefully at the spectra of the guitar's output. Neither of my expectations was confirmed as far as I could tell. So all I can say for now is that different saddle materials seem to alter the tone a bit, but I don't know why or how, exactly. It's probably going to take more time and better equipment than I have to figure that out. Given that the nut material probably has even less of an effect I'm going to pass on that one for a bit.