> It also recouped more than the trial's net cost of 72 million euros ($86 million) through [...] and reduced reliance on other social welfare payments,
Which sounds quite a bit like "we spent more on one type of welfare so we ended up spending less on a different type of welfare." Which, okay, good, but I don't think you can say you "recouped" anything.
If you want to criticize the study, it would be best to actually read the method rather than make assumptions.