Agree that QLN looks dire. 30dB away from a perfect line is a lot of potential lost.
But weirder than that is the relative sync speed, attainable speed and SNR. All happening when G.INP activated. It reminds me of how ECI modems reacted when G.INP was first activated.
Can someone remind me how the HG612 behaves when it doesn't have software compatible with G.INP?