You may want to have a read here:
http://www.aulro.com/afvb/technical-...s-finally.html
But feel free to ignore me. I have only been doing filter R&D for 10 years or so.
Pressure drop across a filter (inversely proportional to airflow) only increases slowly while a filter loads with dust. As noted, those on here who changed a (visually very dusty) filter didn't notice any improvement in fuel economy.
A new filter might remove 90% of the mass of dust reaching it when new, and 99% when dusty, as after a while the captured DUST captures the further influent particles, not the filter fibres. Air filters are NOT sieves. The pores in them are much larger than the particles being captured.
If you read the above tests, you will note that a dusty (donaldson) filter has less pressure drop (flows more air) than an OEM (coopers) filter. So the captured dust has LESS effect than the filter media/construction.

