Bullshit. The touting of hydroxychloroquine did not involve a specific dosage. The studies attempted to evaluate its efficacy and instead found that it made the COVID patient
more likely to die rather than less.
Hospital "compassionate use" involved attempts to use it following all relevant medical recommendations. Notably, hydroxychloroquine is a
very well-documented medication, including for human use (it's legitimately used to treat
malaria,
rheumatoid arthritis,
lupus, and
porphyria cutanea tarda). This meant that information on the dosing and pharmacokinetics for human use was readily available... and hydroxychloroquine has a very narrow
therapeutic window and index, meaning that there's very little difference between a dose so low as to have no effect whatsoever and a dose so high that it'll kill the patient.
This is what your "very specific dose" comment
actually means... and yes, the trials were
very careful to get the dosing within that sweet spot.
"Oh, but they used the wrong [dose/protocol/whatever]!" is a common way for proponents of a bullshit idea to
move the goalposts when evidence is presented against their idea. It's a form of post-hoc
special pleading used to rationalize contrary evidence and allow someone to maintain a belief in the face of overwhelming proof to the contrary -- and the evidence against hydroxychloroquine for COVID has been
overwhelming for years (note the date on that article).
In actuality, the standards for evaluating treatments are generally well-established -- although the debates on the particulars are very much ongoing.
Here is a good resource for learning about them.
And once even basic methodological controls are put into place, the ineffectiveness and harmfulness of hydroxychloroquine for COVID becomes readily apparent. It
doesn't work -- and the side effects appreciably increase the odds of death... and extend hospital stays, etc.