There is a lot of value in learning more about folding (in the gerundive sense), but the focus has always been on finding the folded structure, because as you very well know, structure is function.
It seems the purist definition of solving protein folding means specifying the sequence of actions to arrive at the native structure for a given input sequence. This implies your solution is grounded in a physical reality. Such approaches (i.e MD simulations) have not been very successful to my knowledge and are computationally intractable for most proteins. Even if you are able to do this, how do you verify that your folding mechanism is in fact correct? We don't have ground truth experimental data at the temporal resolution one would want to verify such a prediction.
For now, the only practical framing of the protein folding problem is to predict a native structure given an input sequence. This is simply because we have the requisite data...
The OP is right to complain that there is no scientific knowledge gained by the mere act of modelling a poorly understood process, when the modelling is performed automatically and its results are also poorly understood. The goal of science is to gain knowledge of the processes in nature, not merely to predict their results. That's how science has advanced in the last few thousands of years and it's startling to see Nature and Science standing behind such a different idea of "science".