Capture has always seemed at least plausible to me. The problems with the Big Impact paradigm are many and varied, but most simply is its overuse to explain almost anything that is otherwise a Mystery. Such as the Moon. But also, as examples, the areological dichotomy on Mars, the upside down and very slow rotation of Venus, the on-its-side rotation of Uranus, and at one time, even the presence of the Pacific Ocean bed (which is where the Moon was hypothesized to come from.)

There are other problems. Would large-ish bodies in space actually collide rather than interact gravitationally? The body that is hypothesized to have collided with the Earth to "make" the Moon was supposedly Mars-sized; that's comparatively large. What sort of relative accelerations would be required for these bodies to collide? And how plausible are those speeds in a forming solar system?

We've seen with the repeated impacts on Jupiter that collisions between small bodies and planets are not nearly as rare as was once believed -- which should have an effect on crater-count age dating but probably won't for some time. Yet how commonly have large bodies collided with planets? There's much fretting about how difficult it would be for the Earth to gravitationally capture the Moon (though captured moons seem to be fairly common around other planets), but how likely is a collision on the scale demanded for the Big Impact hypothesis?