Agreed, it seems like the software approach as well. You wouldn't want a class or a piece of code to be "too big" to fail, you'd refactor it into smaller pieces which can be overviewed more easily.
It depends on what you want to optimize for. micro is good for some things, mono is good for other things. Neither is inherently "better", it just depends on your goal
