^ No, I removed the ground from the maf tube as I didn't want to ground the maf in ways it wasn't intended. I'm afraid it might fry something in the ecu and cause a world of diagnostic hurt down the road.
Your corrections in all ranges should be +-5%. That's the common accepted value for a good maf scaling. However, most intakes can not achieve this without additional corrections. This is because the flow pattern past the maf changes with rpm and manifold relative pressure. This is where your MRP correction table comes in. If on flat ground and regular cruise your maf sees 10 g/s and the car is processing 10g/s, once you hit a hill and the manifold vacuum changes for the same rpm, your maf might see 8 g/s and car might be processing 12 g/s leading to a +4 correction. Then in that table you might enter +4 for the rpm and vacuum pressure. So you're making a preemptive correction for the condition then next time you see the same conditions your corrections will be 0 as the +4 g/s needed will come from the MRP table rather then your LTFT. Each intake needs a different scaling for this table as the geometry of the piping varies between intakes.
For my setup, I have this table zeroed out for now. When I tune an intake I first try and get the corrections centered around 0 using the maf scaling, then reduce the corrections using the MRP table. You know something is wrong when you get +-large corrections for the same MRP table cell. That means there is no consistency in the airflow readings for that condition. That means there is a physical problem with your setup, bad sensor, bad signals, or something else.
For my problem, it turned out to be bad signals as well as a bad sensor as far as I can tell so far. The bad signals (maf voltage reading) were coming from the maf sensor using the maf tube as a ground, which grounded intermittently as the engine moved around. Then a bad sensor was causing random large corrections intermittently. Once I ensured the maf has grounded properly and there was no electrical interference, the only thing left that could be causing the problems was bad sensors. So I started swamping out sensors and a new maf seemed to make the difference.
Grounding the maf like you did will take care of the random grounding issue and ensure a clean signal. The only trouble I see with the fix is that the OEM setup was not meant to run like that. Pin 4 on the maf is what gets grounded and the corresponding terminal on the plug does not have a direct route to the chassis ground. It gets grounded through the ecu, which means the ecu is monitoring that pin as a reference signal So if you ground the intake to the chassis directly, you are circumventing this operation without knowing what the end results will be. One thing is certain is that there is voltage running from the ecu, to the intake, then to the chassis ground that is not seen on an oem setup. Since this problem hasn't been investigated before, no one really knows what the consequences of this will be.