Designers who have some knowledge and experience with UX research and testing bring two assets to the table: 1) designs that typically generate fewer negative findings in user test (and, therefore, require less rework afterword); and 2) more rapid understanding of UX research results and how to apply them. (See Henning Fritzenwalder’s original excellent question and the IxDA discussion around this.)
This also works the other direction: UX researcher specialists who have some design knowledge, skills, and experience: 1) find it easier and faster to communicate UX research results to designers (know your audience!), and can often do so through prototype sketches; and 2) are more likely to make user-centered design recommendations that are feasible and cost-effective to produce. The overlap between the two roles and skill sets supports the success of both in designing well for the users, as well as making collaboration more rapid and pleasant between the two.
Can successful test and design happen when the roles are wholly separate? Of course. The distinction can keep each focused on specialties without distraction, and minimize risks of error in research methods and design approaches.
Do I find it more fun and fulfilling when the roles and tasks are circles with a healthy bit of intersection? You bet!