Except in terms of predictive power it’s just barely more successful than random chance. As a discipline it’s still largely in its most primitive stage, and still full of unproven and largely hypothetical assumptions that amount to little more than idle musings and superstitions. It’s so full of holes it’s just slightly more rigorous than pseudoscience. Perhaps in 100 years it will develop into a more respectable discipline, but at this stage it’s at best a tool for short term modelling and post hoc analysis.