[Sesse__]

Postgres isn't really hot on implementing the latest stuff from academia (most of what they do seem to be fairly home-grown, not the least because some of it was never really written about before they started doing it). Arbitrary unnesting is certainly one of the optimizations they don't do; fancy sketches and other estimation techniques generally likewise. But what they're really good at is polish. A thousand small tweaks and optimizations to iron out issues for 0.1% of queries. It tends to have fairly few sharp edges, less so than any other RDBMS I've used.

[fuy]

Thomas Neumann and Alfons Kemper papers can hardly be called "home-grown", their stuff has been implemented in multiple industrial systems. Postgres optimizer, though, is very bad even with simple correlated subqueries, let alone "arbitrary", so it'd be useful to have at least some version of unnesting.