Fair point. I shouldn't have referenced Ethernet in my post and instead mentioned twisted pair cabling.
However, the points I made are still things to consider because they are all related to the electrical characteristics of twisted pair cabling and data transmission over TP cabling.
The protocol used defines how the 1's and 0's are encoded (ie, Ethernet uses Manchester Encoding), for transmission over the physical layer and should be irrelevant. Although, the protocol used is responsible for maintaining the timing between the sender and the receiver, so that the receiver interprets the 1's and 0's at the correct interval, ie. in the middle of the bit time. The Manchester waveform that Ethernet uses is called a self-synchronizing waveform, in other words, there are enough transitions from low to high and high to low that the receiver can stay in synch. with the sender. That's one of the reasons there is a max. distance for Ethernet cables.
It's possible that the protocol being used between the MFC and the Axe-Fx, over the Ethercon cable is having synchronization issues and thus creating the timeouts, which would be compounded with longer cables.
Let me give you an example of how weird, unexpected, things can happen with twisted pair cabling. When I finished school (over 30 years ago now :-( ), I worked for a short time with GE's instrumentation division. My boss was always looking for contracts. So one day he asked us to go to city hall because they were having issues with their mainframe and mainframe cabling. They had hired electricians to pull twisted pair cabling and terminate the serial ends...what a mess. The other issue they had was the mainframe would hang once in a while and they didn't know why. In those days, you accessed the mainframe via dumb terminals, connected via serial cables (RS-232). Basically, if a terminal had data to send, it would assert the RTS signal and the polling mainframe would see the RTS and send a CTS signal. They were short two terminals, so they would leave the cable on the carpet. I suggested scoping the connector and sure enough, the connector, with nothing connected to it was setting the RTS pin high. So the mainframe would send a CTS signal and wait indefinitely. So now it wouldn't poll the other terminals and the system would hang. It didn't always happen and they said it was worse in the winter, so along with some Belden cable engineers, we concluded the RTS signal was going high because of the static electricity created by the carpet, when the humidity was really low in the winter. We told them not to leave the cable end on the carpet and they never had the issue again.
The other thing I learned from the Belden engineers, is that the dyes used to colour the twisted pair wires affects the characteristics of the wire. For example, some colours have better conductivity than others and therefore are more susceptible to cross-talk. So when they twist the pairs, those pairs are kept away from pairs with higher conductivity dyes.
I realize this is not related to the issue mentioned by the OP, but I think it highlights how certain conditions you may not even think of, can affect twisted pair cabling.