I had the same thought! Is this almost-always faster? (If not, would there be a workaround if a user didn't want this new behavior? Postgres core team hates n

I had the same thought! Is this almost-always faster? (If not, would there be a workaround if a user didn't want this new behavior? Postgres core team hates new settings...) Would it be hard it would be to add to the optimizer as a query rewrite?