No - I don’t think that they are particularly correlated. None of my examples above are really “scrambles” - all are across open moorland.
I’m sure there are “scrambles that are pathless” (perhaps the final bit to a summit, where the best way to go varies by season). but there are also scrambles that are extremely well documented as “the best way to go”, but aren’t really paths in the sense that a crap interpreter of OSM data who is creating a “general purpose” map (not a hiking map) would expect.
To be honest, I tend to think that the whole “pathless” discussion is a bit of a diversion from the key problems with path as a tag - it has a ludicrously wide range that requires subtags to interpret sensibly and it even misuses some of those subtags so that you can’t tag legal access properly.