As you pointed out, the viable area with a sufficiently cold winter is probably shrinking every year. Perhaps the better solution is the one proposed by Yablonovitch et al in which they suggest using deserts instead, though I think that is about as risky, and also a tremendous risk for fire.

Deep down, I have the gut feeling that in a century's time, people will think we were clinically insane for not trying to stockpile as much carbon as possible in whatever means we have, since it is clearly the foundation for biology, and widespread synthetic biology will use it (and N, and P) by the gigatonne. I really don't see any fundamental physical reason why we won't be growing entire cities made out of wood with the right genetic engineering and sufficient feedstock.