I meant to point out that, for the use-cases where functional languages are used - production performance seems very acceptable, compared to having to micromanage memory allocation.