All this looks just good, not comments for more than a month, it either means noone is interested, or that noone see any problems in it.
I'm in favor of annoucing it on the talk list, and go to vote, what do you think ? Sletuffe 18:30, 25 March 2009 (UTC)
I approve the idea but i think that natural=mountain_range is more clear. Despite of the fuzziness of this tag, i think that's very important for OSM.
Just an example: Dolomites has been declared natural heritage of the UNESCO and are not present in openstreetmap. We have to think a good way to use Relations and eventually Relations of Relations for sub-groups of mountain ranges. Or is just enough the tag is_in=*?
An example: Sella group is part of Dolomites and Dolomites are part of Alps.
Should we use a relation "Dolomites" to group all the mountain ranges of Dolomites (Sella, Brenta, Sesto, Latemar..)? Or just a way including all them and each of them with the tag is_in=Dolomites ? I don't like this last solution. Firstly the groups of Dolomites are not close each other but sometimes separated.. for example Brenta Group is 50km separated from others but it is considered anyway part of Dolomites. Then is_in=Dolomites is not clear about the exact element which it's referred. Relations are to be used expecially for these things.
Then we could think to group the relation "Dolomites" into a Super-Relation "Alps" together with other groups like the "Mount-Blanc Massif" and others.
The problem with this kind of groups is that we will get a Super-Relation of not necesarely contiguos areas and this is not good because Alps are all together without holes also where there are not real mountain ranges.
The solution could be this: add to the Super-relation "Alps" not only all the mountain ranges like "Dolomites" or "Mount-Blanc Massif" but also an area (element) for the whole Alps.
--Alessioz 15:24, 5 September 2009 (UTC)
I just found this relation that in my opinion is perfect for natural=valley as well as for natural=massif/mountain_range 
--Alessioz 13:00, 6 September 2009 (UTC)
Moving to a relation system using advanced multi-polygons
My comment is also valid for the same looking proposed feature Proposed features/Valley I also think it could be good to merge valley and massif in the more general approach proposition : Relations/Proposed/Region The thing (IMO) we should agree on is the structure of the areas's shapes by using (or not) relation and by the way we add members to the valley, massif, etc. relations and also where do we put the tags.