The linked paper doesn't appear to me to make the claim you did. The combination group (the one from which you could validate the claim you made) is very small - only 486. In fact the paper appears to very carefully outline its own considerable limitations. The authors detail the lack of research in the area and point out a number of conflicting results, including results that directly contradict their own.
That being said, it is a well written, interesting, and thoughtful piece of research. It (and the nearby comment by waheoo) raise quite a few interesting questions for me regarding patient perception and how withdrawal symptoms might interact with marijuana. I also wonder which specific symptoms and underlying conditions might correlate with the various groups. I'm glad you brought this paper to my attention!
Marijuana is a great replacement for those who never started down this path and okay a means of getting off opiates. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6777788/