Anybody remember something similar from a few years back?  (Think BAUT)  I think he is on the right track, BUT that the virtual-particle pairs actually arise in the orientation that requires the least amount of energy, and allows them to persist the longest, in accordance with the HUP.  If my notion is correct, embedding matter in space drives the emergence of gravitation, as a result of the nature of the quantum vacuum.  I don't think that we can expect matter to polarize unoriented virtual particle pairs because of the fleeting nature of their existence.  The pairs must arise in the most advantageous orientation to make this model work.

http://www.physorg.com/news/2011-08-dark-illusion-quantum-vacuum.html

I've often wondered whether it was possible that matter and antimatter were gravitationally repulsive.

Is this beyond our technology now? Is it impossible to get a molecule of antihydrogen in a laser trap and find out whether there is a push or a pull? Is the effect too small to separate from other forces at the atomic level?

If gravitational repulsion between matter and antimatter is not the case, vacuum polarization will no longer be viable, right?

It does seem likely to me that since light is also attracted by matter that the chance of antimatter being repelled by matter would seem minimal. Can't rule it out, though, or some variation whereby the attraction is not as strong, as I've seen posited.

Plenty of proposals out there. Getting the anti-H to low temperatures sounds tough.

Part of CERN's project to produce, capture, and test neutral anti-hydrogen is to find out whether anti-matter behaves differently than matter in gravitational fields.  This level of equivalence has never been tested properly. 

BTW gravitation is a very weak actor, even with very large bodies.  If gravitational repulsion between matter and antimatter is not observed, then we should look to quantum effects in the vacuum to provide polarization.  (See OP).