In my limited knowledge in such things, you have RGB values first, and what follows are the spectral values obtained from the spot measurement. What I do is when the CGATS file contains RGB reference values that had generated the target, I strip out those values and keep the spectral measurements. I then use the result with other software. Even when the RGB values are of the measured target, the spectral values will be more accurate and I think all that is needed for profile generation.
As a side note, I use a UNIX style utility with a GUI front end that understands the extended matching and substitution meta-characters set.
Bob Graham