What I've found is that the best students don't always make the best researchers because of exactly what you've mentioned: there are different skills involved. There are a lot of people who become very good at being a students - who solve any problem set you can come up with and write spectacular reports, etc - but when it comes to the independent thinking and organization required to define a research project, they fall apart.
That said, to be a good researcher you should be at least a reasonable student. It is extremely difficult to go from a state where you stuggle with basic concepts to one where you can make significant progress in an uncharted field.