Exactly. I didn't appreciate this as much as I perhaps should have while working with RethinkDB, but I've been working on optimising SQL queries lately and... man I wish I were working in ReQL. It really made me appreciate how easy (compared to SQL) ReQL is to mentally estimate performance.
With ReQL you new exactly how indexes were being used, and how the data was being accessed. No trying to guess what the query optimizer was doing in the background.