Postgres(PostGis) VS Mongodb: comparing spatial indexes. Who quickly?

Question

PostgreSQL

Colleagues, not for the sake of holivar, but a question of production necessity:
recently I have been working a lot with Postgres and a crazy idea arose to calculate coordinates (spatial indices) in 2D (search for the nearest points, for example a cafe) not to use MongoDB - in which it is now spinning, but postgres. For this, Postgres, as far as I understand, has PostGis. Now the question is - how much will I lose/gain in performance?
Of the benefits: there is one database left (now the main Postgres) and there is no need to upload data to MongoDB.
PS
The fact that Postgis is unrealistically cool and can do a lot - I know this, I'm interested in a simple task - finding the nearest points.

Answer 1

If you already have all the data in postgres, then I see no reason to use Mongo for GIS tasks. Cooler than PostGIS, and now there is nothing free. In terms of speed, there are unlikely to be any differences, indexes, in theory, both systems use similar ones, with the same Big O. On the contrary, you will benefit because there is no need to drive data back and forth.