There's a reason they put the lines in the femoral vessels - these cannulas are like hosepipes - you need to access the large vessels to get sufficient blood flow through the whole circulation.
The amount of flow needed to only sustain someone that hasn't been severely deoxygenated must be a lot less.
A smaller system might be able to initially infuse an oxygen-carrying booster (as Tour de France cyclists do) to get away with lower flow rates.
ECMO not only oxygenates, but pumps to assist the heart. Red blood cells would not resist infusion through small cannulas at high speed. Hemolysis already is a problem with the current size femoral cannulas.
So if what you're after is flow, size is a severe limitation.
I'm aware that's what's taught in med school, though. At the maroscopic level, it works ok.
Severe hypercapnia symptoms are bad.
What would be really amazing is if we could replace defibrillators with these in stores, malls, etc... Instead of trying to restart someone's heart you have a system that's simple enough that a layperson can slap on a dying person to keep them alive while the ambulance arrives. Maybe in the form of a cuff that you clamp over someone's wrist that can use computer vision to identify veins and automatically insert the needles and run the machine. Too complex to be affordable with today's tech, but something that could definitely be possible in my lifetime I think.