Right so this has been nagging at me for a while now. I've been cross-referencing reported Dogman encounters on a couple of the American databases and there does seem to be a clustering pattern around old rail corridors, particularly disused ones. Not just in the States either - there are a handful of UK accounts that follow similar geography.
My working theory, and I'm very much a beginner so take this with salt, is that old railway lines create these long undisturbed corridors through landscape that doesn't get much foot traffic anymore. Whatever these things are, they'd have ideal cover and movement routes. The old cuttings especially, where they've dug down through hills, those would offer concealment and a kind of funnel through terrain.
Anyone else been mapping sightings against historical rail networks? I've started doing it manually with OS maps and printed reports which is slow going but the overlap is interesting enough that I don't think its coincidence. Would love to know if anyone has done proper GIS work on this.