You're both right enough. Aerodynamic diameter doesn't measure the particles themselves, but how their settling velocity compares to a spherical reference ideal of a certain density (1g/cm*3) in a medium.
I don't deal with gas cleaning, but at those scales, if you work a lot with applied processes like filtration and separation, you can ballpark things like daltons with mass and size. I know I do with MWCOs.
Diameter, not weight. PM2.5 is particles of diameter 2.5μm or less.