Sometimes this approach means that you end up with really compact code where you need to do a lot of research and mental modelling to understand if it actually adds up to things being executed in parallel, and if that is good or not.