11

I think that the tools to do the technical stuff already exist, generally speaking. Where there are gaps it’s in the areas of (a) documentation (although https://osm2pgsql.org/ is light-years ahead of what there was before) or (b) “providing a web front-end as a service for people to do things with”.

With regard to (b) anything too broad in coverage would be well out of scope for a GSoC project. What might be useful would be finding a subset of “interesting problems” that can’t be answered by an Overpass query and which aren’t already documented examples from either the osm2pgsql or postgresql / postgis documentation. The sort of things I’ve had a need for have included checking for path breakage and multipolygon breakage.