Yeah this has actually caught my attention too. There's something to this pattern I reckon. Old rail lines tend to follow the most ancient routes through landscapes - valleys, ridges, natural corridors that humans and animals have used for thousands of years. Makes sense that something territorial would stick to those same corridors.
We don't have much Dogman activity over here in Yorkshire but I've been cross-referencing American sighting reports for about two years now and the rail line clustering is hard to dismiss. A lot of the Midwest cases especially seem to sit right on old freight lines or abandoned ones.
My theory, and its just a theory, is that the rail cuts create these long straight channels through woodland that act almost like highways. Whatever these things are they seem to use them for movement rather than territory marking.
Anyone mapped this properly on GIS or anything? Would love to see an actual overlay of sighting coordinates against historical rail routes. Feels like the kind of data that could actually mean something if done right.