I'm not positive exactly how this factors in, but it's well established in the oilfield industry that pressure can have a very dramatic influence on "permeability" (approximately rate of diffusion) of gases in whole rocks. Measured perms in "tight gas sands" are frequently lower by a factor of 1000 at realistic native confining pressures than at atmospheric pressure. I would think that there should be some sort of analogy between this inter-grain permeability and the diffusivity along crystal defects that would control helium flow out of zircons. I'm not too sure I have the perseverance to look for documentation, though.