Test with full street network

Until now, we have tested the corridor and segment delineation with streets down to the "tertiary" level in the OSM hierarchy. We can test how the result differ if more (all?) levels are included. Apart from performance, are the methods built robust enough to avoid too much segmentation of the blocks?

Start by locating the corridor and segment delineation methods and their current tertiary-level street-network input. Compare results when additional OSM hierarchy levels are included, checking runtime and block segmentation; done means documenting whether the methods remain robust and how performance changes.

Written by the indexing model from the issue text.