The LF trigger mode is working correctly.
When you try to trigger on 2 alternating traces, the signal going to the trigger circuit includes a lower freq square wave, which is a composite of the trace position of the 2 signals.
This lower freq square wave gets differentiated by the AC coupling to the trigger circuit.
So, the trigger sees a differentiated spike with the vertical signals riding on the differentiated slopes.
Note that the trigger instability improves when you position the 2 traces closer together.
This is because the "square wave" component of the trigger signal is being reduced in amplitude.
Trigger off either of the signals & the problem will go away.
You could also try DC coupling but you may still have the problem.
"Chop" mode usually does not display this problem because the chop rate is much faster than the Alt rate.
So, it doesn't get differentiated by the AC coupling to the trigger circuit.
The purpose of LF Reject triggering is to address this specific problem.
You will not find LF trigger mode on a single-trace scope.