Psst: a real science decides between competing hypotheses or models with experiments. There are no decisive experiments to test the big disagreements in economics because the economic issues are riddled with historical context that can't be disentangled. The data are historical so the factors can't be isolated: they're subject to interpretation in emphasizing some factors and ignoring others based on political ideology. These are evaluative questions about how society should be organized to achieve certain goals. The comparison between economics and physics is a grotesque distraction that's had social Darwinian implications, tainting the progressive modern project.

Libertarian economic theory, for example, is unfalsifiable. It doesn't matter whether the model seems to fail to describe how things work, because the model is prescriptive: it sets up a goal to be reached, and if society fails to achieve it, to maximize individual liberties, you can always double down on the prescription and pray that the market will work itself out in the long run.

Likewise, progressive economics is unfalsifiable. If government intervention (the picking of winners and losers) impedes innovation or makes for certain market inefficiencies, that's a tradeoff these economists are willing to make because of their values or priorities.

These aren't descriptive, scientific questions of the matching of models with facts. The economic models are prescriptions or creeds disguised as rigorous descriptions of natural systems.