I seem to follow a pattern with most web 2.0 tools. There's an initiation phase where I try out the tool and I use it for a short amount of time. In that initiation phase, I figure out how it works, but I'm only scratching the surface of what the tool can do and I'm already noticing some of the drawbacks.
Then my attention span drops off and I barely visit that tool for months at a time. My suspicion is that this is where most people give up on a tool and decide it's not for them. I've done that recently with Quora.
At some point, I come across something on the web that reminds me that I have an account on that tool and I revisit it. Very often, the tool has evolved and added new functionalities between my first and second trials. I'm usually happy with improvements and likely to pick it up again. The second trial tends to be more focused on getting something specific out of it... a more focused project. It doesn't imply that I'm going to use the tool on an ongoing basis, just that I've thought about what the tool can be useful for and when I might need it, not necessarily on a daily basis. This pattern was realized with Pearltrees. I played with it more than a year ago, found it somewhat interesting but limited in the way it structures links between pearls (much less flexible than a mindmap for example, yet much easier to create than a mindmap consisting only of URLs).
One of the drawbacks of Pearltrees is that a Pearl has to be a URL. As far as I can tell, you can't add a "concept" pearl without a link. It's not meant to build concept maps or mind maps. If I'm organizing links, after the first ten links, I'm automatically starting to think about how to group them around key concepts. Pearltrees doesn't allow you to do that easily. To address that challenge, I've used links to Wikipedia as a way of organizing around key concepts in the pearl map below.