The cross reactivity data is mostly from animals and the current version of fujian-like flu has already drifted from last years A/fujian/411/2002 which I believe is what was used in the animal tests.
The next flu batch (for southern hemisphere) will probably use A/Wyoming/3/2003, which is from a Feb 2003 isolate, and the current version has already drifted from Wyoming also.
WHO is clearly playing "catch up", but may be accelerating mutations by using a vaccine that is only partially effective against Fujian (although 20-30% of H3N2 in the US is much closer to the Panama strain in the current vaccine).