As to answer the survey, I use wifi on my N1, N2, and N2 Plus, as well as my custom printer based on another N1 control box and mainboard.
The hypothesis is that the PHY and actual physical routing of traces to the Ethernet connector- specific to the N series custom IMX6 board in the Raise Touch control panel has some level of interference or noise, and when running at maximum rate, this noise then causes errors in packets and chokes the system.
To that end, yes, I personally have seen this in testing of my 4 N series printers on Ethernet. I saw various slowdowns and slow tranfer rates, regardless of raisetouch software version up until the most recent 0.9.10 which has rate limiting function to avoid this. Also, in my fleet of various printers starting from the earliest kickstarter unit to the most recent shipped N2Plus and a replacement panel for that same N2Plus, I have various versions of the IMX6 board (early boards were white, later boards are blue and a different design) and ALL of them show some form of slow rates compared to using the wifi interface on my network.
However, I also have a data point that I doubt anyone else has. That is, I wanted to run the Raise 3D interface on other printers, and since Raise 3D will not sell that part of the electronics. So, I worked and worked to get the Raisetouch complete software and operating system image copied over to a genuine Linksprite brand pcDuino Acadia IMX6 quad board. It also uses the exact same Ethernet PHY chip, the same processor and so forth, but is just a slightly different board. It running the exact same code, does not appear to have this problem on the Ethernet.
Again, the same code, same processor, same PHY Ethernet IC, however minor differences in the actual board itself between the PHY chip and the physical Ethernet jack. And yes, this produced a data point with a different result.
Going further, and why I said we need to make is clear on this survey and post- Pro2 owners have a different and completely new IMX6 board completely different with all new layout. We absolutely should collect the data points of if these folks use wifi or ethernet, but we need to make that distinction is we are trying to compile this survey into meaningful data, rather than just lumping them in with N series users since the board and electronics are in fact different.