It's not difficult to construct a situation where dipole attraction would be unstable, but gravity isn't. E.g. an odd number of point particles evenly spaced in

It's not difficult to construct a situation where dipole attraction would be unstable, but gravity isn't. E.g. an odd number of point particles evenly spaced in a circle. (Real answer: plausible-sounding explanations are ten a penny. Which is why scientific theories claiming Y is an example of X generally need to either (a) mathematically derive the governing equations of Y starting from those of X, or else (b) make a testable prediction about Y that would not be true if Y was not an example of X).