Interesting one this. I've mainly been into crop circles and poltergeist stuff so Dogman isn't really my area, but the railroad clustering idea caught my eye because it mirrors something people have noted with other phenomena.
Old railway lines follow natural contours, they cut through woodland, avoid settlements, run near rivers and embankments. So you'd expect any large cryptid to use them as natural corridors the same way deer and foxes do with old drove roads round here in Somerset. Could be mundane wildlife behaviour rather than anything stranger.
That said I do wonder whether there's a genuine data bias here - people are more likely to be out walking old trackbeds, so naturally more sightings get reported from those locations. Has anyone actually mapped this properly against population density and footfall? Because without that baseline the clustering doesn't really mean much does it. Would be good to see some actual numbers before jumping to conclusions about ley lines or whatever the current theory is.