Been thinking about this for a while actually. There's a well-documented black dog route across the northern moor here that follows what historical maps show as a medieval parish boundary almost exactly. I've been out with a camera at those crossings maybe a dozen times over the years and the cold spots, the odd light anomalies in long exposures, they cluster right at the boundary markers.
My working theory is that these old boundaries weren't arbitrary. They followed older land divisions, possibly pre-Christian, and something about liminal geography concentrates whatever phenomenon is producing the sightings. Black dogs specifically seem tied to liminal spaces - crossroads, county lines, parish edges.
Has anyone done a proper overlay of UK black dog sighting databases against historical tithe maps or enclosure award documents? I'd love to compare notes. The Ordnance Survey historical mapping archive is genuinely useful for this kind of thing if you haven't looked into it. Would be interesting to see whether this pattern holds up outside the West Country or if its more regional than I'm assuming.