I've seen some people do that with internal filters and power heads, but for most of those setups, the air tube input is positioned just before the filters outflow, so the Co2 would tend to just flow out in large bubbles and then float quickly to the surface, hence not having much contact time with the water to dissolve properly.
I guess if the Co2 input is put at the filter's inflow, then the Co2 can get chopped up into tinier bubbles as it flow through the sponges and filter media and diffuse into the water better, though that might create air pockets which could affect the impeller system.
Bookmarks