I've seen this argument before, and I think the real place it falls down is that there just isn't that much space. For mature growth forest, you have something like 40-ish big hardwood trees per acre [1]. Those are, to my understanding, the only ones that are really relevant because they make up most of the biomass. Other types of trees are just a rounding error. `1.2e12 / 40 acres -> square miles` (frink) is 4.7e7 square miles. That is more than the whole land area of Earth [2]. There is certainly not enough land that would support mature forests available to plant so many trees, unless I have made a serious calculation error.
[1]: http://www.sbcounty.gov/calmast/sbc/html/healthy_forest.asp
At any rate, if that's what the research is getting at, they probably need to title it differently, or explain early on that merely planting the trees is a necessary but not sufficient condition to achieve the wins they envision.
