Re Somalia / Somaililand (trying to wrap things up in the absence of further comment from the person who thought OSM was wrong):
What OSM has are areas that are essentially the maximal “claim” for both Somalia (as admin_level=2) and Somaliland (as admin_level=3). This is fairly in keeping with other places around the world. It’s true here (as elsewhere) that both entities’ control over those areas is weak. There are other lower-level entities there (e.g. Puntland) that as I understand it don’t claim statehood, so Puntland in OSM is probably also OK, even though it appears to be in dispute with Somaliland. Essentially then what OSM has is probably “as good as it can be” within OSM’s data model.
Of course, if someone has other information (and not just from the usual news sources, wikipedia and the CIA World Factbook) then please do say so - my knowledge of the situation in this part of Africa is remote and likely to remain so.
