Herbert, the OEM Hyundai part in US retail around 150 USD (on ebay) while the OEM part for BMW is more than 300USD.
I think Jarance's graph look ok because the linear increase of the air mass vs RPM (constant load in this case). I think our MAF is hot-wire type where the wire is heated to some 300 degrees F, when air passes through and cools the wire the MAF will attempt to keep the wire at that same temp, the amount of current needed to keep the temp is proportional to the air mass entering the throttle body. That is why when the MAF is dirty the cooling is not so consistent. Not sure about BMW MAF, most hot-wire MAF will heat the wire to a very high temp for a short while to burn off the contaminants when turn off the engine.
My FC is not that good because more than 70% City use. I can get less than 10L/100KM for highway at cruising speed. I have done many things, vanos seals, check the DISA, check leaks, and I will clean out the vanos solenoids and the little piston/valve soon. I always suspect the O2 sensors are sluggish, because I feel sluggish O2 sensors will affect the most stop/go and city type of driving where RPM/speed keep flunctuating. The O2 sensors voltage should flunctuate at about .45 volt at fast as possible so air/fuel mixture can be optimized. All these are based on my limited research. I am still trying to use GT1 or other software to see the frequencies of the O2 sensors, before I change them out.
I have read that some have tried driving with MAF disconnected, if it car perform better with MAF disconnected then the MAF is bad. I think the logic here is that , the DME will operate in open-loop with preloaded map when the MAF is disconnected so the performance should be better.
These are strictly my opinion.