I guess you used max mode for both the v12 and the omniglide : the lower airflow of the omniglide should be the reason why the cylone dust seperation is not working that well compared to the v12. But since your test is simulating the dust amount of a longer period is should not be such a huge problem.