Great points.

For isolated projects in the past I’ve had great success storing things in postgres,. (generally large documents that are relationally tied to other more traditional PG data, in a db driven project) Just saying size and recall have been pretty happy.

As you say the other features are distinguishing.