Cloud Functions (PHP, Python) hold onto idle connections to Cloud SQL despite max pool set to 1

Question

My Cloud SQL instance is showing many more connections than I would expect:

I have a python cloud function that I've set the following properties for when connecting to the Postgres Cloud Sql instance:

connection = sqlalchemy.create_engine(
        # Equivalent URL:
        # postgresql+pg8000://<db_user>:<db_pass>@/<db_name>
        #                         ?unix_sock=<socket_path>/<cloud_sql_instance_name>/.s.PGSQL.5432
        sqlalchemy.engine.url.URL.create(
            drivername="postgresql+pg8000",
            username=db_user,
            password=db_pass,
            database=db_name,
            query={
                "unix_sock": "{}/{}/.s.PGSQL.5432".format(
                db_socket_dir,  # e.g. "/cloudsql"
                instance_connection_name)  # i.e "<PROJECT-NAME>:<INSTANCE-REGION>:<INSTANCE-NAME>"
            }
        ),
        # Pool size is the maximum number of permanent connections to keep.
        pool_size=1,
        # Temporarily exceeds the set pool_size if no connections are available.
        max_overflow=0,
        # echo=True
    )

I believe since the pool size is set to 1, the # of connections on the Cloud SQL side, should be roughly the same as the cloud function instance count:

My two other cloud functions are PHP which according to these two stack overflow posts: Connection pooling in PHP How to establish PHP PDO connection with pooling

don't have connection pooling.

The Instance Count charts for these look like the following:

In total, I would expect a max connection count of around 20 but it is consistently getting up to 50+, so I'm looking for any ideas on other causes.