For top pairs with invariant mass above 450 GeV the asymmetry is claimed by CDF to be stunningly large 48+/-11 per cent. 3 times more often top quarks produced in qqbar annihilation prefer to move in the direction of q. If true this would favor color octet excitations of Z0as the most natural explanation since the asymmetry would be not only due to the interference of vector and axial vector exchanges but also due to the inherent parity breaking of colored Z0 couplings. The effect would provide further support for the identification of color quantum numbers in terms of color partial waves rather than as spin like quantum numbers. The earlier support comes from the evidence for colored excitations of leptons.
Addition: After a badly slept night I have come to new thoughts about the possible explanation of the effect. What is so weird (really weird when one begins to think the numbers!) that the outgoing top quark (t) remembers the direction of motion of quark q before annihilation to intermediate gluon which it should by the basic definition of annihilation diagram. For any exchange diagram the situation would be totally different: consider only Coulomb scattering! The quark q of the first proton would scatter from the quark of the second proton and transform to top quark in the scattering and keep its direction of motion in good approximation since small angle exchanges dominate due to the propagator factor. Flavor changing exchange diagrams are however not possible in the standard model world since the only flavor changing are charged weak currents and their contribution is negligible.
In the new physics inspired by TGD situation is however different! The identification of family replication phenomenon in terms of genus of the wormhole throats (see this) predicts that family replication corresponds to a dynamical SU(3) symmetry (having nothing to do with color SU(3)or Gell-Mann's SU(3)) with gauge bosons belonging to the octet and singlet representations. Ordinary gauge bosons would correspond besides the familar singlet representation also to exotic octet representation for which the exchanges induce neutral flavor changing currents in the case of gluons and neutral weak bosons and charge changing ones in the case of charged gauge bosons. The exchanges of the octet representation for gluons would explain the anomaly! Also electroweak octet could of of course contribute.
What is fantastic is that LHC will soon allow to decide whether this explanation is correct!
Addition: I noticed that this argument requires a more detailed explanation for what happens in the exchange of gauge boson changing the genus. Particles correspond to wormhole contacts. For topologically condensed fermions the genus of the second throat is that of sphere created when the fermionic CP2 vacuum extremal touches background space sheet. For bosons both wormhole throats are dynamical and the topologiies of both throats matter. The exchange diagram corresponds to a situation in which g=gi fermionic wormhole throat from past turns back in time spending some time as second throat of virtual boson wormhole contact and g=gf from future turns back in time and defines the second throat of virtual boson wormhole contact. The turning corresponds to gauge boson exchange represented by a wormhole contact with g=gi and g=gf wormhole throats. Ordinary gauge bosons are quantum superpositions of (g,g) pairs transforming as SU(3) singlets and SU(3) charged octet bosons are of pairs (g1,g2) g1≠ g2. In the absence of topological mixing of fermions inducing CKM mixing the exchange is possible only between fermions of same generation. The mixing is however large and changes the situation.
Addition: The following sayings by some celebrities of science deserve also to be added because the last three decades in theoretical particle physics have convincingly demonstrated their truth.
"The difference between stupidity and genius is that genius has its limit."
"Only two things are infinite: the universe and human stupidity, and I`m not sure about the universe."