Take the above example - we really don't know how it works. We measure the amount of dopamine/serotonin in the spine before and after you take it so we assume it increases the presence of those chemicals in the brain, but it's not really mapped out.
A lot of people have the misconception that we understand the body the way we understand other technologies. That you can balance a chemical equation and deduce how something will effect the body.
While we understand a lot, this really isn't the case. And for all practical purposes it sometimes is treated like "magic".
You know what they call alternative medicine that works?
Medicine.
(cf: Tim Minchin, 'Storm' http://www.youtube.com/watch?v=HhGuXCuDb1U)
One reason is that it takes a lot of modeling to understand why to dose 1x twice a week vs 2x once a week, so everyone gets roughly the same dose.