The process of diagnosis when it involves actual testing is legitimate. Which does include ADHD. Provided your comfortable gatekeeping it behind only people who can pay hundreds of dollars for neuropsyche testing.
But the process where it concerns anxiety, depression and bipolar? It's not. These do not have any biomarkers in diagnosis, and psychiatrists are not actually experts at identifying this stuff by asking a handful of questions for 15 minutes. Their years of training doesn't make them have some magical ability to identify a soup of random incoherent symptoms accurately, and they are not significantly more accurate than moderately well informed patients at identifying themselves. That's why people are so commonly misdiagnosed. It's literally just the person's opinion. And you shop around for the right opinion. Then they actually diagnosis you with whatever is required to get the insurance to pay for the medicine. That's what diagnosis actually is. It's a paper to have insurance pay for medical care. And it is not science.
Some books on the subject if you want more authoritative sourcing:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4817669/
https://www.bps.org.uk/member-networks/division-clinical-psychology/power-threat-meaning-framework
I think we mostly agree, given you acknowledge misdiagnosis is rampant within psychiatry. Is it rather that you see the DSM itself as legitimate, and doctor are just misunderstanding it which causes misdiagnosis?