Correct. In absolute reliability of the system during high round count testing, no CRF has been able to match the best PF for MRBS. However, that doesn’t mean they aren’t very reliable, just that they don’t have the same MRBS in high round count mil and government testing. Understanding however, that isn’t including human errors causing malfunctions.
An odd school AI is probably the all around most reliable bolt action made- though the Parker Hale M85 might argue; however, they don’t make a 6lb rifle.
The reason PF has a higher MRBS is simple- push bolt forward, round chambers. It’s way easier to get a PF manufactured correctly. CRF requires more work to get timed correctly and in general CRF actions don’t do well when push feeding common DBM’s.
However, a correctly built CRF is extremely reliable, can work with push feeding from a DBM, allows the user absolute control of ejection (with manual ejector), and removes the greatest single user caused malfunction- short stroking causing a quasi “double feed”.
Given ideal rifles in both- PF has less mechanical malfunctions, CRF has less human induced malfunctions. Given well designed and correctly built versions of each, it’s doubtful one could see a functional mechanical difference…. But you will see a difference in user induced malfunctions.