I believe what was meant was that assuming local minima of a sufficient size to capture your probe, given a sufficiently high density of those, you become extremely likely to get stuck. A counterpoint regarding dimensionality is made by the comment adjacent to yours.