See also this article by Steve Silberman.

One problem I had when reading Silberman is that there was some vagueness around what the "placebo effect" was supposed to be. He was constantly referring to it as if it were a real thing, but that struck me as a figure-ground inversion: a controlled experiment doesn't "measure the placebo effect", it tries to accurately measure the effect of the drug's active ingredient by keeping every other effect the same (i.e. identically distributed). (Things cleared up a little for me when I read Peter Lipson's response to Silberman, though his critique also has some vagueness to it.)

So what "the placebo effect" really means is "every random effect other than that of the active ingredient". It does not mean "the beneficial effects of taking a pill that you think is going to help even though it contains no active ingredient", even though that's the meaning implied by these articles.

The linked article is different in an interesting way. The study referenced measured the difference between a "no treatment" group (i.e. no pill but the same interaction with doctors) and a group receiving treatment consisting of an avowedly inactive pill patients were told would help through "psychological effects".

One problem I can see is that the instrument used to measure "effects" of the "placebo" is itself psychological: the IBS-GIS as far as I can tell is a questionnaire given to the patients. So quite possibly this study is replicating priming effects, not necessarily measuring actual difference in health outcomes.