It tries to draw conclusions about the relative divergence times of different genes based on degrees of sequence similarity. This evidence is extremely weak, as there are much more likely explanations for the observed patterns of sequence variation, such as a shift in the selection pressure on a gene down the chimp/human lineage, for instance. Despite this weakness, they chose to highlight the sensational conclusion that human and chimp ancestors must have swapped genes through mating long after the presumed speciation time. This was irresponsible, because it caused a widespread public sensation when the paper came out.
it seems pretty harsh to lump it together with [1]...
I'm not lumping them together. The common characteristic is that they're both garbage. The conclusion of [2] is a subtler form of it but not to an evolutionary scientist: the problem with it was immediately obvious to me and many others as soon as it came out. And it's a more serious corruption of scientific discourse because it's a Nature article. The data the analysis was based on has merit, though.