The studies cited by this metastudy all suffer from the same issue: They aren't studying the cannabis plant at all and even if they are, it's not in the chemovars (chemical makeups) that real consumers are actually consuming - due to cannabis sourcing issues stemming from cannabis's continued illegality. For randomized controlled trials, even in "legal" states, university scientists can't just walk into a dispensary…
>The headline could read instead: No evidence cannabinoid isolates help anxiety, depression, or PTSD. There’s no evidence that what they tested with was pure THC isolates. If they’re using cannabis in plant form, even if it was bred for higher thc content, there is still cbd.
tldr; "If they're using cannabis in plant form" is a very, very high bar for the current state of cannabis (really cannabinoid) research.