Conversely a club-mate had a move misreported when he slid a rook too slowly on an e-board and it registered one square too soon, losing the game. His carelessness of course. Not a chessnut but another board/interface where he couldn't set a latency.
My beef is quite simply, they know that some e-boards (at least, the chessnut air) triggers something in their cheat detection software, they cannot reasonably interpret this as cheating. Because they know it's not a signature of cheating, it's just a signature of using an e-board. Regardless of approval or whatever. It's not cheating and wrong of them to label it as such.
I can perhaps see one potential problem with Chessconnect. It allows you to set your own latency. What if I set a really long latency, like 2 seconds or longer? Would it allow me to make a move, even briefly take my hand off the piece, see immediately that it's a bad move and move it back, before the move even registered with chess.com? That would seem like cheating to me if it works like that, though I admit I don't really understand latency!
Latency has always been an issue even before these e-boards became useable. There was even a time when latency was exploited to cheat the clock from running out back in the early 2010s.