But an analogous situation holds for almost all performance benchmarks of any product right? That seems to be an argument against ever making any benchmarks public.