>That just can't be right. Great SQL people understand set theory, query planning, and a lot of arcane RDBMS internals that most of us don't have to get into.
Your likely misinterpreting what the author means with great.
If your not at a scale where a team is pushing the micro optimization boundary, there is tremendous value in the ability to keep everything simple. Having the entire org operate using the same frame is great.
What's more, it's not uncommon for great simplicity to be mistaken for 'obvious', and excessive complexity be mistaken for great work.
Which is also why ORM are so popular despite being suboptimum. Not because they save you from writting SQL, but because they force people to standardize : to define a single point of entry for they data, to centralize the schema, integraty checks or signals, and of course, normalize the way the query are written.
This has greater value than perf for a lot of orgs.
Your likely misinterpreting what the author means with great.
If your not at a scale where a team is pushing the micro optimization boundary, there is tremendous value in the ability to keep everything simple. Having the entire org operate using the same frame is great.
What's more, it's not uncommon for great simplicity to be mistaken for 'obvious', and excessive complexity be mistaken for great work.