therapeutic effect as well in a "chronic condition characterized by fatigue, widespread pain in the muscles, ligaments and tendons, and multiple tender points — places on your body where slight pressure causes pain."
Almost anyone who has had a massage when their muscles are sore knows that it can relieve muscle pain; I haven't had acupuncture but others have told me it can relieve pain as well. Pain can cause you to hold your body abnormally which can cause additional muscle & joint pain.
There are some quite a few poorly controlled studies published looking at acupuncture, but I would expect the Mayo study to be done on a higher level. Since I can't read the article (it's in process at PubMed; link below), I don't know exactly how their placebo acupuncture was done, but usually it is done by touching the skin with the needle but not inserting it in the same way that is done in acupuncture. You should read more studies on acupuncture before rejecting any one study out of hand based on a few paragraphs in a summary article intended for a general, non-specialist reader. One wouldn't expect such a summary article to give in=depth detail about the procedure - the original article is the place to look for such detail.
50 patients, split between two groups, is not necessarily a tiny sample size - it all depends on the effect size of the treatment. The researchers probably began with a pilot study and ran a power analysis on those results to determine what size of sample should be adequate to provide statistically significant results. They then would run that size of a group and if the results weren't significant, they'd be out of luck. Additionally, no study is going to simply trust respondents to self-diagnose; it goes without saying (and will be detailed in the journal article) that the physicians themselves determined that the respondents indeed had fibromyalgia. This is standard in medical research.
One can certainly control for the degree to which an observed effect of treatment is due to the placebo effect: this is the purpose of the placebo group. If the placebo group shows some improvement, that is the "placebo effect". If the treatment group shows greater improvement than the placebo group, you can establish that while part of the treatment group's improvement is due to the placebo effect (as shown with the placebo group), since the improvement in the treatment group is greater than that for the placebo group, part of this improvement has to be due to the treatment administered. Of course, you run statistics, and depending on what kind of statistics you run and how stringent you are, you can say that with X% certainty (where X is usually between 95 and <100), your results are not due to chance.
Important for any study is the characteristics of the group being studied. One wouldn't expect the results of the study to extend beyond people who don't have approximately the same characteristics. Also, the treatment can be considered "effective" as long as it is effective in enough patients in the "treatment" group, compared to the "placebo" group, to give statistically significant results. In other words, an "effective" treatment does not have to work in everyone. If it's not, the next step is trying to characterize what factors might differentiate those in whom it works vs. those in whom it doesn't work (a lot of prescription drugs aren't effective in every patient - this is why doctors try one for a while and if it's not giving the desired results or has too many side effects, they try another. In medicine, the next giant step will be developing tests (blood, genetic, etc.) which will be more predictive about whether a medicine will work or not).
Finally, few physicians would advocate a therapy like acupuncture to be the sole treatment approach if there are proven drugs available precisely because there hasn't been much well designed research looking at acupuncture (but the body of well-designed studies in the literature is growing - well-designed studies require funding, which is only now becoming available, albeit in relatively tiny amounts). More likely they would approve acupuncture for an additional treatment method; acupuncture could, for example, reduce the amount of pain medication a patient requires.
The article isn't available yet at pubmed but here is the reference; you can also directly contact an author and ask for a copy of their article, although if you go to any major university library, they're likely to have (in=library for general public) access to the journal (you can call and ask):
Martin DP, Sletten CD, Williams BA, Berger IH.
Improvement in fibromyalgia symptoms with acupuncture: results of a randomized controlled trial.
Mayo Clin Proc. 2006 Jun;81 (6) : 749-57.
PMID: 16770975