In order to expect data sharing, you have to be open to collaboration, yes? Just wanted to point that out. Science is ruthless in its own way, or at least, the scientists and publishers make it that way.
I can be completely opposed to collaboration and still expect data sharing. This does not necessarily mean that I will get it. It also doesn’t mean that I won’t. Now to address what you think you said: it depends on what data you want to share, and how open everyone is with it, depending on what the value of the information is to each person. If I found some interesting new metamorphic structure somewhere, there is no way that I would be able to publish on it or even collaborate, simply because I don’t have the background to deal with the technical side of things past an elementary level. But someone else could. There is no need to sit on something you find out of jealousy if NO ONE is ever going to get a paper out of it because of you.