That is all policies and political structures are testable and tested to see their effect on those three (or other suggested) factors. If a policy doesn’t reasonably work then it’s simply not continually employed. I’m curious to see what factors others think ought to be used.
It seems most political systems now were built without science in mind and utilize it as an afterthought to help develop legitimacy for policies individuals want. Generally politics across countries seems deeply emotionally driven and not fact driven. That is people have a feeling that an idea is a good idea and then they cobble together whatever they can to support that point without any unified measure of good or better. Ideally it ought to be the other way around, fact or evidence informed policy generation.
I don’t think they’re talking about p hacking. P hacking is when the statistician keeps adjusting the test to get a significant result. What they meant is probably Goodhart’s Law.