It's worth reading the author's actual peer reviewed paper: http://www.nature.com/articles/ncomms2380
> We estimate that free-ranging domestic cats kill 1.3–4.0 billion birds and 6.3–22.3 billion mammals annually. Un-owned cats, as opposed to owned pets, cause the majority of this mortality. Our findings suggest that free-ranging cats cause substantially greater wildlife mortality than previously thought and are likely the single greatest source of anthropogenic mortality for US birds and mammals.
Other parts of the world have had wild cat populations for millions of years.
The US has a few native cats, and I assume has had them for quite a while.
While the models they use are fairly conservative, it seems like figure 2 is pretty telling. 50% of the variance in their estimates comes from the "Unowned cats" parameter - a parameter with which "no empirically derived data exist".
IMO think they are reaching too far on this - but it is a nice elegant, simple study. I look forward to following the letters re: this pub.