I love this library but it often takes a while to run. Could we speed it up by running all (or at least most) of the tasks in parallel?