Well there is a reason why doctors kept telling me I am a hypochondriac, but I do have a whole zoo of conditions simultaneously, and this is a pretty common state for people with hEDS and I'm on the extreme end of it. So while milder versions of it are ~2% of the population the extreme versions of it are < 1/20K.
And yeah, medical researchers are in fact in the aggregate really bad at their jobs. Look how long it took to convince surgeons to wash their hands. But a lot of the genetic stuff relies on Linear Regression for GWAS which assumes independents of SNPs, otherwise you get multicollinearity problems, this is not a safe assumption and they've confused their results as confirming their assumptions. Instead of listing everything they get wrong a much shorter list is what they get right, Dr Jessica Eccles (https://x.com/BendyBrain) does great research into Long Covid and Generalized Joint Hypermobility which should put to bed the theory that GJH is benign - still good luck trying to talk a doctor out of that train of thought.
If you find someone who has hEDS the odds are they have a very large number of those things and most of them don't even know the names of most of the conditions, just one or two that bother them the most. The RCCX / hEDS list is a distinct subset of all possible things, the list of all medical maladies is far longer. It becomes highly improbable that a set of people have the same set of maladies - doctors tend to chalk this up to social contagion but that doesn't bear out. Genetic and behavioral causes have distinct diffusion patterns.
It's confirmable with WGS which I've done and I've encouraged many others to do and it turns out that you can indeed predict with a great deal of reliability if someone has TNXB / CYP21A2 SNPs. Unfortunately it's harder to find people who have C4 since they're likely to have schizophrenia.
I have an Ancestry partial genome that I've imputed to expand. Would be curious whether those SNP's are present in the data.