> scala is real big impediment to making data processing accessible to general public in your company

Ding Ding Ding! Presto/Athena now is becoming huge in BI ecosystem. We don't really use Spark for ad-hoc BI anymore, we use it for data science and large repetitive workload.