This and other threads on ramp testing is a good read.
The major issue with a ramp test (and any test that relies on your ability to perform effort above ftp to determine ftp) is that everyone’s ability above ftp is not identical and so everyone will have a slightly different result. Those right in the middle will get a probably accurate result but everyone else will not. You also run into issues where you can do some anaerobic work and improve your ramp test result but not actually improve your ftp.
TRs ai ftp and workout feedback seems to do a solid job at pushing people closer to an actually accurate ftp.
I am not going to assume anything on how tr ran the data for their ramp test but I don’t think they ever released any data comparing ramp test results to ftp.