So "mass" isn't "mass"
it depends on "other factors".
This disparity seems to suggest it's not only the
mass flow through the sensor but the
flow's trajectory through the sensor, i.e., the flow can be non-uniform flow across the aperture of the MAF sensor and the sensor is actually only measuring the flow in a certain subset of its aperture.
Am I surprised? I guess not, there's only a few wires in the sensor and if the flow's non-uniform then... "of course".
Did I expect this? No. I guess I hadn't thought about it.
So anybody who sticks a non-OE style of intake / air box in front of the MAF, or relocates the MAF, may require a "retuning" of the OE MAF flow curves.