Yep, that dang Rover.
One might well view Lakatoss MRP as a synthesis of what was acceptable in Popper and Kuhn. His main point is that, contrary to naive falsificationism (i.e., Popper), theories of a certain sort the sort that are cores of research programmes are not sharply falsifiable. They can be cumulatively disconfirmed over a period of time, but they can't be decisively knocked out by a single crucial experiment.
This point is charmingly illustrated by the imaginary case of planetary misbehaviour which Lakatos recounts in Falsification and the methodology of scientific research programmes. But in truth we should not need a detour through case-studies, real or imaginary, in order to grasp this point which is that deep scientific theories are no more falsifiable than they are verifiable.
http://www.shef.ac.uk/~phil/courses/312/05lakatos.htm