No, that's not at all what this is about. A better one-sentence summary would be "Accurate data can be very misleading if, for example, it's viewed at the wrong level of granularity."
Well, that seems a particularly tricky chart to interpret.
In this case, I haven't read the article the chart was taken from; I don't know what argument the chart is supposed to support. Stripped of that context, it's a pretty confusing chart. It seems the author (of the SciAm article - Cairo) is using the chart to make a point about lying with charts. I don't think that Cairo is publishing his research on obesity and birth-weight - that seems to be Kitahara et al. If that's what's going on, then it's hardly surprising that the chart is hard to read; Cairo chose it to make exactly that point.
And I think he's being unfair to Kitahara et al., implying that they've deliberately contrived that chart to mislead.
The article is more about how you can read the graph as it is intended (showing a positive association, for example) but reading it uncritically means that you will ignore the possible context that generated the statistics.