I will give your request a shot. It is only a theory. The theory is the earlier code was faulty and transmitters were running at full power on all antennae regardless of heating, impedance, standing wave ratios and whatever else etc. Lets say they updated the code to fix that oversight that the FCC might fine them on. Very likely scenario on a new product rushed out the door.
An RF engineer will tell you that you just cannot throw in a random cable into the middle of a tuned array that depends on identical antennae and impedance to work. All antenna must be tuned identically to the exact wavelength in use or there will be signal loss (or if the system is smart enough it will shut down the transmitter to avoid damage if the SWR is severe enough). Nor can you remove one antenna in a tuned system and expect it to keep working, that is another type of impedance mismatch and possible source of RF interference that would run afoul of regulations.
In the meantime you have created a kluge with a 4 meter extension cable on one RF port introducing an impedance mismatch of biblical proportions versus other antennae and other issues on a router with the old code running. Yes it sort of worked with the old code but probably not as the FCC requires. Now with the corrected code the system correctly senses an impedance mismatch on the 2.4 Ghz antennas and backs off on the output power to avoid overheating, failure, spurious out of band transmissions, warranty claim, lawsuits and whatever else.
The correct solution would be to run a 4m ethernet cable or POE cable from an asus lan port to your shed and use an old router as a switch/AP rather than as a router. Turn off the old router's DHCP, set a local ip address and plug the cable at the shed end into a LAN port not a wan port on the old router.
If you still want to pursue this adventure with the RF cable it would be helpful to pull out an antenna design book and calculate the exact number of full wavelengths at the middle of the 2.4 GHz band on the cable to the shed and cut the cable to that exact length. At these frequencies mm matter. In addition to increase the probability of this working I am going to say that you need to have exactly identical extension cables on ALL 2.4 GHz antennae to keep the system balanced, but my guess is that the mere addition of an extension will mess up some fancy advanced RF reception feature never mind the transmitter.
Or you could just run the old code.