It's not the first piece of research or the first article suggesting this, though so far I've only seen this kind of thing in relation to SSRIs.
Personally, I'm not bothered if something works due to placebo effect, provided it's not causing significant side effects, I'd be bothered if a doctor told me I needed to stop breastfeeding to take a medication which it turned out was probably only helping for that reason.
However, with depression (and probably other illnesses too), if you take part in a double blind study and you don't get better, or get worse, you get dropped from the study because regardless of whether you're getting the drug or the placebo you need something different, so by the end of the study, there is a risk you end up comparing two groups that are no longer matched.
There is definitely a place for such studies, but it's all about how you interpret the results, because "no better than placebo" doesn't mean encourage doctors not to prescribe, it's got to be "no better than doing nothing" for that to be the recommendation.
My PPD has been severe (as it was the other time), so according to these studies SSRIs should work, but depending on which one I've tried it's either made no difference, or even made things worse. It's frustrating for me that SSRIs are so popular that most doctors I've worked with spend a long time increasing doses or switching to another SSRI when another antidepressant would be better.