But in so doing, it would largely cease being a science, because--so far, at least--we have only modest reason to believe that "systematically articulating the conceptual structures" actually gets at anything that a good novel does not, and with far less entertainment value and memorability.

There may not be any there there, in the majority of cases.

It is in uncovering causal mechanisms or at least sufficiently strong correlations that they have predictive value that psychology finds itself as a science.

And yes, alas, the systems are not simple. We can approach an electrically-isolated vacuum if we need to isolate a system to reduce confounding variables to conduct experiments in physics, but the socially isolated human is far outside their normal behavioral regime and we're intrinsically reactive and responsive anyway. The physics of keys absent locks is not very interesting.

Systematic articulation of conceptual structures makes the same mistake that psychology has always made: it wants to start at the top but say things with a special kind of confidence that you can only do when your ideas are tested thoroughly by experiment and observation, and the tests are, for the most part, not possible.

But at the bottom, if one is humble, the tests are possible. We can assemble a vast wisdom of little anecdotes, find curiosities like lack of transitivity in comparisons, and see whether from our bag of tricks and confidently-made observations patterns begin to emerge. Even if they don't, they can still have practical benefit: cognitive behavioral therapy is built in large part off of research into the fundamentals of habit formation, and CBT and its friends are, while usually not transformative, pretty consistently helpful.

Starting the other way, at the top, with such heady concepts as "conceptual structures that shape human experience", just invites confirmation bias--as psychologists should be all too aware--as we take our extremely imperfect theory and gather supporting evidence because to try a disproof would be pointless: at all times there are mountains of contradictory evidence.

This isn't to say it shouldn't be done at all. There are usually a few low-hanging fruit on any investigatory tree. Just that this is unlikely to be the path by which psychology develops as a robust, deep, reproducible branch of human knowledge.