Robot with "human-like" vision

With robots being outfitted with many different types of sensors, it's not surprising that human researchers are interested in replicating how we perceive the outside world.  This article from Anne-Marie Corley looks at a new robot that uses stereoscopic vision in order to assess the the robot's location and the speed that it is travelling. It replaces a more traditional method of robot navigation using laser range finders, and interestingly, the article states that it is probably not the most efficient method for robot travel.

It is easy to see that it is not the practical application of getting a robot to do a certain specific task, but more research into how the human brain works.  Ultimately the funding to do this type of research comes from a completely different pot than industrial robotics research and development, but one can't rule out this technology finding a practical application sometime in the future.

In your opinion is this type of research worthwhile?  Or should research money be used in a more practical way--for solving a specific real-world problem?