Possibilities:

1) The Flynn Effect is based on bad data.
2) The Flynn Effect tracks non-g rises in IQ.
3) The Flynn Effect measures a rise in g. Therefore better environments improve g a lot.
4) The entire concept of g is somehow faulty
5) The entire concept of IQ is somehow faulty.

I think that we can rule out 1 and 5 easily enough. I don't think that 'g' is sacred. I can certainly imagine a new theory supplanting it, but on the balance, I think that we can say that 4 is improbable.

I have heard both 2 and 3. Rushton, for example, claims 2, while Flynn claims 3. I'm willing to believe Flynn because he's more mainstream.

So that leaves 3. So now for the million-dollar question, what about the environment is improving IQ so much?