You might be surprised at the number of medications which are effective for specific conditions or symptoms but for which we don't know the mechanism that achieves the effect.

Yes, modern medicine has a large component of throwing stuff at the wall to see if it sticks.

If it sticks, and it passes all the usual clinical trials to prove it does more good than harm, it's approved, even if we don't understand exactly why it works.

Try looking deeply into almost any specific condition - let's take heart disease. Cholesterol, for instance - is it bad? Is it necessary? Do you need a certain amount or does the amount not matter so long as you have the correct ratio of fats? Why are some people perfectly healthy with cholesterol numbers off the roof, and others can have atherosclerosis with cholesterol numbers off the charts low? Will reducing cholesterol for that second person help them? Hurt them?

For the record, the treatment plan my cardiologist put me on was explained as "all these things seem correlated with not having heart attacks so we'll just do them all and hope it helps you".

I think the best the pill can do right now is to at least making some disturbance to change the balance and hoping it produce desirable outcome.

Yes basically throw something at the wall and hoping it stick.