OpenStreet map is increasingly becoming an intermediary between us and the real world. We use a version of simplified reality of points, lines and relationships processed by volunteers. We ask a rhetorical question. Given the increasing density of information, could OSM be used for scientific research? Of course, this is happening, but could it be possible to ensure that at least small, compact areas - national parks, some specific spaces closed by borders - receive information as accurate as possible thanks to OSM? We tried it and we want to share some observations on this subject.