Any chance that your benchmarks Michael could have helped identify the performance regression? If not, why not?