[ENH] Add workflow for processing and adding new subjects to BABS results

[mattcieslak]

Summary

BABS currently is set up to run an entire BIDS dataset and combine its results. This is great for datasets that have finished collecting data, but this is rarely the case.

Next steps

We're currently working on a situation where FairlyBig was run on most of a dataset and the result was pushed to AWS/github. Now we'd like to start with the dataset from github and run more subjects. I don't know what the most efficient workflow would be for starting FairlyBig with a dataset that already contains a bunch of results.

Has anyone tried this yet?