That's because economics is a social science, and you never know exactly how people are going to react. A study of one culture might not apply to another, and a study in one timeframe might not apply to another.
So we will probably see a lot of "hmmm... that's interesting" kinds of things that might not be repeatable. And lots of good and bad effects may be unmeasured or unmeasurable.
Doesn't mean we should stop doing drug trials, or economic trials in this case. Well designed experiments can be informative, even though care must be taken to generalize them.
I work in pharma (biologicals) - though not as a scientist - and you're overstating the differences. The genetic variance of humans is within a very narrow band. The superficial differences are not relevant to the vast majority of pharmaceutical products. In any case, how our human biology reacts to pharmaceutical substances is VASTLY different from humans' economic and social interactions. You know to a very high degree of certainty how an antigen will produce an immune response much much more than how pricing your headphones at 50 dollars will get you more customers, or raising wages by 10% will give you a 20% increase in productivity.
>Every science faces an enormous amount of confounding variables.
I don't think (macro)economics/human psychology are sciences. Aside from my own opinion, it is certainly debatable - Unlike biology and physics which are firmly in the science camp.
If somebody announced a drug test that didn't look at symptoms or benchmark against placebo effects was "testing a new cure", people would rightly point out it was doing no such thing, just as an evaluation of how people in extreme poverty react to charitable handouts doesn't really tell us much about the effects of redesigning the Western welfare state on UBI principle that taxpayers really ought to subsidise wealthy early retirees and housewives on exactly the same basis they subsidise the desperately seeking work and disabled.