There are lies, damn lies and statistics. This article suggests that teams stop throwing the ball to their backs because it's not "efficient." It's stuff like this that gives sports stats such a bad name. Like Bastiat's bad economist, the bad statistician focuses solely on what is directly in front of them rather than looking at the big picture. It starts with the obvious fact that throwing to WRs (mostly downfield) is better than throwing to RBs, mostly short check downs and often behind the LOS. The authors then beat the dead horse with statistical proofs of the obvious.
Assuming a competent QB, the throws will be mostly to open receivers so of course going to an open WR downfield ia going to yield better results than going to a RB short on average. And the real question is should the QB throw to a covered WR rather than an open RB? Secondarily does it make the offense better to take away a threat of passing to the backs? They don't address those questions at all.
The focus on early downs is particularly disingenuous. Threat of the run will keep the LBs near the LOS and create more room behind them. That definitely favors WR patterns over those typically used by RBs in the crowded short zones that are even more crowded on 1st down. Does anyone really need to be told that a team with Ezekiel Elliot might be better off throwing a PA pass over the LBs than a swing pass when the defense has 9-10 guys line up within 5 yds of the LOS? His other example just reinforce the bias. Neither the NY Giants nor the NE Patriots are teams that spread it out and throw the ball around on early downs.
I'd be much more interested to see what the numbers look like for pass heavy teams running spread attacks. I'd bet those numbers look a lot different for the KC Chiefs for example. I would also love to see how the productivity of the offense generally and of the other receivers specifically is affected by incorporating the RBs into the passing game. I have no idea where to even begin with that one though.