Since the LXRHash is fairly solid now, I wanted to decrease the amount of data collected. The benchmark utility now only runs two times, once for a single-thread performance, once for multi-threaded performance (using the number of available cores).
Running a benchmark should now complete in a little over 2 minutes.