Manually run benchmarks to identify performance problems with Bonsai. This task is focused on benchmarking the optimizer itself, and not the quality of produced plans.

Initially the task will use the benchmarks from the mongo-perf test suite.