I think labs that are running unreliable or outright false tests need to be stopped. However, I can see potential problems even with this as there could be turf wars or people trying to stop competitors, so there would need to be independence in the regulatory body. At the very least the test centres should have to write on their test results that they have been deemed unreliable, if they are running dodgey tests.
The treatment thing is trickier I think, as sometimes advances in medicine comes from trying things out. Unfortunately harm can also come via this route. I would be more hesitant about shutting down treatments for this reason.
But tests should be doing what they claim to do and this is more measurable. I think also outside calibration might be necessary to avoid either outright scammers or cock-ups. There was a bit of a scandal in my local hospital when it was found that people were given false negatives on results. S0 screw-ups can happen even in more established institutions, if someone incompetent is employed.