What is the reason that the docs recommend using a separate postgres server instead of the one deployed with the helm chart?

Databases can be tricky in the kubernetes world of abstractions. To generalize, you either have the ops expertise/team that knows exactly how you want to manage your DBs, or you don’t and are likely best off with a fully managed offering. https://cloud.google.com/blog/products/databases/to-run-or-not-to-run-a-database-on-kubernetes-what-to-consider

Interesting article, thanks!