pg_net: Add max_concurrency parameter (or include concurrent requests in the batch_size)

As far as I understand, the batch_size it is just the amount of request being dispatched every tick but this does not take into account running requests. My current use case is that once a record is inserted in a media table, I call a function to process the image. It is indeed going into a queue which eventually processes everything however in some cases I want to limit how many concurrent requests are happening at the same time.