Seeing as 90% of all medical research is flawed, (and no, this is not some crank - it's pretty much accepted in research because of the evidence, as well as researchers personal knowledge/experience of bad studies we need to test against placebos, because some of the supposedly beneficial treatments are later found to be harmful.
Remember thalidomide? Or more recently, the panic over hormone replacement therapy because the biggest, best trial of HRT ever was started prematurely because it purported to show harmful effects of estrogen? 10s of millions of women world-wide were suddenly put on antidepressants to help deal with menopause side-effects. Turns out that the study was bogus, but more than a decade later, many doctors still haven't got the memo.
The flaw in the study was in using only estrogen from pregnant mare urine (Premarin and Prempro). Equine estrogen is not bio-identical to human estrogen (estradiol estrogen is), but it also contains equine enzymes that the human body has never seen in nature and can't handle - which cause, among other things, liver failure. Having horse enzymes in your blood, your organs, your brain
Heart attack is the #1 killer of women (no, not breast cancer). Estrogen helps protect against cardiovascular diseases. So people on a placebo would have been healthier, with less chance of sudden death.
Studies also show it's a potent anti-depressive, and delays the onset of Alzheimers and other dementias. Those are sentences of a slow death. It also reduces or stops suicidal ideation.
It also slows down bone demineralization by enabling the digestive system to take up more calcium (which is why calcium supplementation by itself doesn't work - if your body can't absorb it, you'll just eliminate it). 28% of women and 37% of men who fracture their hip die within one year. Considering that 1 in 3 women and 1 in 5 men will get at least 1 hip fracture, that's a significant amount of women who are now at higher risk because of this "greatest study ever."
So even the best studies need to be re-done, because if this study had included a placebo as well as Premarin, and also included other sources of estrogen such as estradiol, Premarin would have been singled out as the biggest contribution to health risk from HRT.
So placebos have a place - they would have saved millions of women from premature death, not endangered them further. That's why we do studies, and why we include placebos. Plus, placebos also work even when the patient is told that it's just a sugar pill. That's why you compare the benefits of a course of treatment with a placebo as well as no treatment. Why prescribe a drug with bad side effects when a placebo performs either as well or better? Not including placebos places people depending on the results of the study to make informed decisions at risk of making bad decisions. Some of those are fatal.