I too have the mysterious dotted automation lines appearing willy-nilly well past the end of the song, and I've not been able to discern exactly why they get put there. Was it something I inadvertently caused to happen? I do a lot of cutting and pasting during arrange sessions and I suspect it has something to do with that but I don't know for sure. I would love to know what the "rules" are for these things.
As to the end-of-song marker, it would be useful for me. I often record bits in a non-linear way and place them well after the body of the song, then use them when I'm ready. I often end up not using these bits in the song I'm working on for whatever reason, but like them enough to not want to lose them. Incidentally, I'd love to have the ability to put these in some kind of "idea bag" in a way that's well integrated in the workflow.
Anyway, I often only want to render the tune without the music blocks at the end, and an end-of-song marker would be useful to easily exclude them. Yes I know I can highlight the timeline to render only the parts I want, but that often results in many more clicks because I tend to render a lot to test the sound quality at various mp3 quality levels. Also, I'm often zoomed in, so to select the timeline part with any accuracy I usually have to adjust the zoom level first, then select the timeline portion. Not a big deal, but quite a few clicks that an end-of-song marker would tend to substantially reduce.
All that being said, I don't consider an end-of-song marker to be a high priority, must-have feature whereas I do consider the anomalous automation event issue to be important.