Regarding the exploit (4A Gigabit Global Edition), I also had the same behavior of needing a reboot after setting the password, for it to work.
Another thing.
I know that weaker wifi on this unit on openwrt in terms of range is probably caused by the change from proprietary drivers to the opensource ones.
Searching a bit I found an interesting thread regarding a newifi3 d2 and signal range
...an excerpt
The problem with Ralink/Mediatek devices is the poorly calibrated EEPROM on many devices.
On my Newifi D2 I tried changing the factory partition which contains the EEPROM, wireless performance immediately went up without even needing to change the wireless signal strength.
https://github.com/gwlim/art-radio/tree/master/RAMIPS-Newifi-D3-Factory (same chipset and radios as 4a gigabit)
I wonder if we used calibration data from a different mt7621a device that has the same radios, maybe something could be optimized ? (under the assumption that we also have suboptimal calibration data from the beginning)
Does anyone have some research on this topic ?