As I mentioned, I have a spectrophotometer, and I could do the spectral measurements if someone wants to supply the appropriately exposed and developed film. If anyone is interested then let me know.
To be meaningful the test would have to satisfy certain requirements. For example, shooting random scenes would not work.
Here are a few minimum requirements. At least Two rolls would need to be supplied, each roll would need to be exposed under identical conditions. Frames would need to be exposed in solid colors (absolutely no details present in the images). Several different colors would need presented, each in a different image. The colors should have good spectral coverage, i.e. a frames in pure red, a pure green, a pure blue, and maybe some other mixed colors. (I am over-simplifying a little bit here.) Roll 1 would need to be processed in conventional C-41 by a highly reliable lab that uses only the best quality consumables and practices. Roll 2 would need to be processed using a C-41 kit that uses blix using only freshly prepared chemicals whithout re-using solutions.
As I mentioned, this is a minimum requirement. A better program would be that the rolls be grouped into three groups. Every roll would need to be exposed identically, as indicated above. In each of the three groups there would need to be at least three rolls, and in each group each roll would need to be treated absolutely identically. (The purpose of this is a replication study.) Group 1 would be rolls that have undergone conventional C-41 processing. Group 2 would be rolls that have undergone processing with a C-41 kit that uses blix, as described above. Group 3 would be similar to group 2 except that the film would undergo two separate blix steps. The first step would be using once-used blix (but otherwise fresh). The second step blix step would be using absolutely fresh unused blix. Ideally, between the two blix steps there would be a water wash. This is to simulate what would happen if one were to run a process in an automated machine that normally uses separate bleach an fix, but with blix substituted for both the bleach step and the fix step.
This scheme would be used to check for similarity of color produced by the different processes. It would not evaluate which color is "best", and it would not evaluate other factors, such as grain or long term image stability or other factors.
It's probably feasible to evaluate grain by a slight extension of this scheme, but it might be more expensive because it might require drum scans, although a high quality prosumer scanner such as a Nikon or Minolta (or even a Canon) might work. However, it would require someone to become involved who has enough image processing experience to be able to extract some statistical information, i.e. the pixel-to-pixel standard deviation of the images.
If anyone is interested then we can discuss it further, including a more detailed/stringent specification for the study.