I would guess that there is a bias towards the direction the debris was orbiting. But thinking about it a little more, initially, there are no Lagrange points as the debris is spread out in a ring, then the Lagrange points develop as the moonlets become massive, so they could get trapped that way, rather than needing a braking mechanism.
I've got an article describing a simulation I ran (
http://www.orbitsimulator.com/gravit...onbuilder.html ), where 100 moonlets combined to form the Moon. Although in this simulation I didn't get any Lagrangian moons, other simulations I've ran did produce Lagrangian trios. These simulations are over-simplified, but they can serve to show what's possible. So maybe we once did have Lagrangian moons.
Simulations of the current Earth / Moon system show that objects placed on the L4 and L5 points are perturbed away by the Sun, almost immediately. I think my record for trapping one is about 20 years, but usually they last only a few months to low years. So as a trio of Moons spiraled out, at some point their Lagrangian configuration would be destroyed, leaving us with only 1 moon.