Yeah this has been on my radar for a while actually. Rail lines follow old geographic corridors - valleys, ridgelines, river paths - that were already established travel routes long before trains existed. So you're not really looking at a rail connection, you're looking at a landscape corridor connection. The rail just happens to sit on top of it.
There's a cluster of Texas sightings I've been mapping for about two years now and three of them fall within half a mile of an old Southern Pacific route that got decommissioned in the 80s. Thats not nothing. Old rail beds also tend to be heavily overgrown now and relatively undisturbed by development which makes them decent habitat corridors for something thats trying to move without being seen.
What I'd want to know is whether anyone has cross-referenced this with pre-rail indigenous trail networks because my suspicion is the overlap would be significant. These things aren't following the rail lines, they're following the same routes everything has always followed. Would be good to get some proper GIS mapping going on this rather than just anecdotal clustering.