Author of trailrouter.com here. Your analysis sounds correct - I am indeed limited by the source data available, and you correctly infer that I rely on "feature

Author of trailrouter.com here. Your analysis sounds correct - I am indeed limited by the source data available, and you correctly infer that I rely on "features" available in openstreetmap (rivers, parks, etc). Nonetheless, would you mind sharing your location and I can take a look anyway? Feel free to email directly, my email is on the website.