It is good to know of real-world misdiagnoses made via 23andMe.
Now, can we see some real-world statistics by the government-licensed medical profession? I wonder what a side-by-side analysis of the errors 23andMe has made vs. those the medical community has made might look like. I wonder how the real-life consequences might stack up.
As far as I know, 23andMe has never accidentally amputated the wrong limb, so it's at least go that going for it.
Interesting tidbit: you can always double-check the reports 23andMe gives to you by getting followup tests someplace else.