JSONB data type, ALTER SYSTEM statement for changing config values, ability to refresh materialized views without blocking reads, dynamic registration/start/stop of background worker processes, Logical Decoding API, GiN index improvements, Linux huge page support, database cache reloading via pg_prewarm UPSERT, row level security, TABLESAMPLE, CUBE/ROLLUP, GROUPING SETS, and new BRIN index[113] Parallel query support, PostgreSQL foreign data wrapper (FDW) improvements with sort/join pushdown, multiple synchronous standbys, faster vacuuming of large table Logical replication, declarative table partitioning, improved query parallelism
We'd all love for software to be bug-free and stable, but features have to be added.
It also depends on what kind of software we are talking about. A trendy mobile app is very different from a critical database.
I personally prefer software that’s more conservative in its approach to new features. It’s becoming harder to find software that has this philosophy but not yet impossible.
I agree with you: all features have to have a cost and value. The market decides what features are valuable, and they get built and supported. SELECT EXCEPT is not one that the market values yet. And that's fine.
I'm curious as to what you envision as "the market" in the context of free software, or Postgres in particular, since I don't expect it's the cliche of people literally deciding which product to pay for among competing ones.
A couple possibilities came to mind, but, having no intimate familiarity with any major project like this, I didn't want to parade mere assumptions as multiple choice.
If skipping EXCEPT means say the postgres team can focus that much more energy on bugs and stability, then that's a good trade off in my eyes.