From my experience, in any dissemination activity you should first think about two things:
1) Who is your core target audience?
2) Where does your core target audience search for datasets/what is the main community "hangout" in that respect?
This will give you a good indication where to put your data - and once you have figured that out, you can worry about DOIs, metadata and all the rest.
This post has been migrated from the Open Science private beta at StackExchange (A51.SE)