Adding the clinical framing, because it changes how the question reads. The honest answer is that the effect is real, the magnitude is contested, and the individual variation is larger than either. Those three things can all be true at once, and most arguments here are two people holding different parts of that.

kate.chem said:

hank_denver said: ...compounded vs brand cost and coverage...

There is a second half to this that has not been said yet. It helps to ask what evidence would change your mind before you look at any. If nothing would, the discussion is not about evidence, and it is better to say so early than to spend nine posts discovering it.

The figures, for anyone assembling their own picture. For anyone reading later: the numbers in this thread are worth checking against a primary source before you act on them, including mine. Half the figures circulating in this community trace back to a secondary summary that dropped a qualifier.

One thing that is still open after kate.chem’s answer:

How long did you give it before you decided it was working?

Moderator note: leaving this open. It is being argued well and the disagreement is the useful part.