The Capture vs. Retrieval Problem
Any note taking system (or perhaps any knowledge system, including project management applications) have to solve for two problem: information capture and retrieval. If information is not easy to capture, people will not add information to the system. If information is hard to retrieve, people will not be able to get utility out of the system.
It's not so much a versus as it is a tightly coupled problem. Sometimes one is lost at the expense of another. Sometimes at both. For example if people never can't retrieve information, the capture problem becomes more difficult, as people ask, "why bother?"
It's a common problem in software documentation. Developers have tried to solve the capture have by auto-generating documentation. However, this fails to capture more abstract information that needs to be written in different forms so that it can be understood by different people.
Capture vs. Retrieval also relates to how and when information is structured. One criticism of systems like Evernote is that you need to define the structure before you can start capturing information. This violates galls Law: By trying to design the system upfront, you will inevitably create an overly complex system that does not work and requires refactoring.
Some systems subvert this by giving you an "inbox." The getting things done system is one; knowledge base software like Roam is another. Both give you a place to put data in an unstructured way. Roam attempts to make it easier to retrieve by encouraging links and creating a self-referential graph. Getting things done solves by adding a third step, a processing and organization page. It makes capture easier by removing organization from part of the process, and makes retrieval easier because you will have a system intentionally designed after information was collected.