Yeah, I figured you know, it was meant for people who didn't.
Is it necessary to get to the molecular level of how the placebo effect works in that specific case though to use it? Just like we don't know with many medications, I think it is not. If it's one of the two
1. natural healing
2. mind-body (neurobiological) interaction leading to either actual healing or reduction of symptoms
Shouldn't we be able to use both without figuring out the exact pathways? Clearly there are limits to what a person is prepared and able to do and it might be easier or better in some cases to give/take a pill or whatever other placebo (which is why I wrote the above caveat), but in general I think the better way is to rely on treatments that are effective in themselves whenever possible (meaning having a greater effect than placebo alone).
Sure, placebo effect is part of all treatment. While I agree that you cannot separate out the effects of placebo and the treatment itself in the individual, you can on the larger scale in studies: give a treatment to one group and a placebo to the other and if the groups are large enough you do get a good grasp on the part that's placebo and the part of the effect that's caused by the treatment.