> Yeah that's my point! That's the load per instance that I see at my current company, we just have fewer replicas.

Yup. At a previous company and current, I had single instances handling 120K QPS.

> If you get it wrong (and getting it right is not trivial), production is on fire and it's potentially hours of downtime (or days, if it's a massive table!).

You’re not wrong. Hopefully stage is representative enough to gain confidence. For self-hosted, I use the indisnotvalid method, but I do get that it’s not a feature per se.