Even just learning how to use CTEs effectively has been a major improvement for my daily work. Building up a dataset iteratively in clearly defined steps and performing aggregations, transformations, orderings, etc along the way is just a very clean way to work. And it's easy to short-circuit at any step for debugging if you're not getting the results you expect. I often work with datasets larger than the amount of memory on my computer so has benefits there as well instead of trying to load it all into Python memory to work with it.

[0] https://www.postgresql.org/docs/9.1/queries-with.html