Inverse: pedestrian shortcuts

[dabreegster]

https://www.openstreetmap.org/way/671146174 as a motivating example.

Look for a driving route between the endpoints of a footpath, and calculate the detour ratio. High values are better pedestrian shortcuts. (Complications include having to take into account all the complexities of driver routing, like turn restrictions and which direction to do the routing. And even in this example, both endpoints of the footpath aren't driveable -- have to search further on the graph.

An analogous route / score mode would be helpful. And showing both scores in one place could be interesting, to find areas favoring one mode.

[dabreegster]

Closer to this, but it's weird in a few ways.

This anecdote isn't showing up as dramatically, because I think https://www.openstreetmap.org/way/83592298 is missing restrictions.

Anyway,

• Think through "mode comparison" approaches. For an individual route, compare directness. In score mode, have a way to look for places where detour factor is very different for two modes.