i have always thought that the air filter was one of those things you can't measure on the dyno, if it adds any hp it will be so minor that it will be well within any standard deviation of the dyno tests.

maybe with all out FI set up there would be a measurable diference between filters. on most set ups i think it is a decision between the best filter(paper) or the best flow(k&n type) and in either case there probably wont be enuff diference to matter.