My part of political science has been (attempting to) follow scientific method since the 1940s. Its results have been... underwhelming. The main reason for this, bluntly, is that people are really damned complicated. They're difficult to model, they behave irrationally, even a minimal model has way too many variables. Hell, they don't even vote using the same heuristics that they did 50 years ago. There's a reason why abstract social science (like model-based economics) can look like it has little connection to reality, while the less abstract you get then the woolier and less scientifically satisfying it generally becomes. It's because getting crisp, reproducible, cumulative, results is extraordinarily hard.
Faced with that, you have two choices: keep trying to get better at what we're doing, or decide that these questions aren't worth answering (or that it isn't possible to). I think that better understanding the behaviour of people collectively organising themselves is vital, so I pursue the first. But I have sympathy with those that would prefer to work on something else.
Right, the scientific method is mostly useless for social science which is why it is in such a dire state. My point is that we need something of equal significance for social science before it will start make noticeable progress, until then it will be on the level as Aristotelian physics.
Social sciences have their own methods, which are suited for the unique kinds of data and questions they deal with. Read John Law’s After Method: Mess in Social Science Research for a solid overview of why you can’t just mask natural science methods onto the social sciences, and which might also clue you into why your assumptions regarding the value of social science research are wrong.
And yet the Chinese are successfully using real actual Scientific Socialism ("with Chinese characteristics" - but Scientific Socialism nonetheless!) to raise hundreds of millions of their fellow countrymen from the most extreme sort of poverty. You don't get any less "woolly and unsatisfying" than that!
Faced with that, you have two choices: keep trying to get better at what we're doing, or decide that these questions aren't worth answering (or that it isn't possible to). I think that better understanding the behaviour of people collectively organising themselves is vital, so I pursue the first. But I have sympathy with those that would prefer to work on something else.