> The title does not reveal that the suggestion applies only to Postgres.

I don't think the OP intends the suggestion to apply only to Postgres:

> The name-giving PostgreSQL database is often already available and in systems where it is not, *it can be easily introduced with minimal effort.*

… i.e., don't have PG? Not a problem, just introduce it into your architecture! (/s, from me, but I think the OP is serious.)

> It is a bad idea to tie your architecture to a specific technology.

But I agree. These sort of "just keep wiring boxes together until it works" architectures cause so many headaches IME; so many pieces, each with its own failure modes and bugs. This is where I find the claim the OP makes,

> [this design, with PostgREST] is more flexible than one that is developed manually.

… dubious. No, the advantage of doing it in code I control is that I control t hat code, and if something needs to change, it can be. With wired-together-boxes-of-random-tooling like this, I have to pray my change fits into the whims of its config files.