I am interested did you figured out how to make this work!? I know its an old post but I find this usefull.
the method he applied is sound. might suggest a few more flow rates to make sure he had a nice smooth curve but otherwise it should work.
some more information about how far things were off might help.
I just did this recently with 2013 sp1.0 and it worked as expected. the only thing that might cause an issue is if the calibration density was checked. that might artificially scale the values. but that would only be if a density other than what was measured was either used or entered.