The article seems to be all over the place and mixes interesting phenomena like human biases and fallacies with pure speculation, like claiming "bots can easily ruin social networks" which they infer from merely running simulations. Obviously their simulation doesn't prove anything about the nature of networks of real humans (as one example).

Afaik their Botometer tool is also highly disputed, supposedly having way too many false positives.

They also use supposed "independent fact checkers" for reference, when those "independent fact checkers" on Facebook and Twitter are decidedly left leaning. This in turn casts doubt on their conclusions on the differences between Democrats and Republicans (for example something like Republicans are more likely to believe fake news as verified by independent fact checkers - or maybe left leaning fact checkers are more likely to label "conservative news" as fake news).

Ironically, the whole article is too long to comment on it all, kind of supporting the title of it (information overload helps spread fake news).

I find it very concerning how people start to overly rely on "studies". It really needs to become common knowledge that a "study of this" or a "study of that" really doesn't prove anything just yet. Those studies need to be scrutinized and verified and replicated, and properly interpreted.