To me the biggest advantage is expression. You can package up a whole traversal or other optics into a single expression and then you can give it a name, refer to it, augment it. For example what if later you'd like to add another `if` into the inner loop? Would you duplicate the code? Would you add a callback to the original imperative code? The big idea here is that access to some subpart of a big object (like a path) can be abstracted into its own expression type and made composable, and the less important idea is that it can also be well-typed.

The second advantage is conciseness. I don't know if you've actually used jq, but if you have, I really don't know why anyone prefers a verbose for-loop-and-if-statement block over a single concise jq expression. Paul Graham expressed this idea better than I could have so read this: http://www.paulgraham.com/power.html