We're definitely going to revisit this periodically and I would love to address this in the next iteration. Would you mind creating an issue at https://github.c

We're definitely going to revisit this periodically and I would love to address this in the next iteration. Would you mind creating an issue at https://github.com/fivetran/benchmark https://github.com/fivetran/benchmark describing the concurrency trade-off that we're not accurately capturing?