We should collect some kind of timing data as well as scientific data for comparison of the new vs old schemes.
It looks like CLUBB is compatible/instrumented with GPTL, so should be a case of extending this and gathering data over controlled extended runs.