I think it would be useful to be able fold the protein in stages. This is easiest described by an example. Imagine a protein that is 200 amino acids long. Instead of folding the whole thing at once, I'd like to be able to fold it in stages. So for example I could get the first 50 amino acids, fold them, add on the next 50 amino acids, fold the resulting 100 amino acid protein and so on.

This has some advantages:

i) it would enable me to focus on a piece of the protein without being distracted by other parts of the protein
ii) it mimics somewhat the process that goes on in cells - presumably some proteins start to fold while the are still being constructed by the ribosome (I imagine there are also mechanisms that, for some proteins, stop them from folding until they are fully constructed, so this staged approach probably won't work for those proteins)

The number 50 is just an example - the number of amino acids to add needs to be selectable by the user - part of the skill will be in deciding how many amino acids to add at each stage.

Of course piecewise folding won't give an optimal folding for all proteins, but I still think it would be a useful tool for some proteins.