A key year in their analysis is 2008 when "the average length of the golf course on the PGA Tour increased by 37.4 yards". From that year there is a dramatic increase in the returns to driving length and accuracy.
To illustrate their argument on earning they compare the 2013 earnings of Bubba Watson and Bryce Molder. Watson was the 5th longest driver and 122nd best putter. Molder drove the ball 14 years less than the average golfer but was the 4th best putter. Watson earned $1.8m while Molder earned $0.8m. Eight of the top 10 drivers earned more than $1m whereas only 5 of the top 10 putter earned over $1m.
Carson, Day & Burford suggest that the old saying "drive for show and putt for dough" should be replaced by "drive for dough in order to putt for dough".
The evidence presented in the paper is not incompatible with Rotella's views. However, it clearly attaches a greater importance to driving than one might get from Golf Is Not A Game of Perfect. One of the things I found most interesting about the paper is my reaction to it. If someone asked me to guess which of two golfers earned more money, and they said one was a better driver while the other was better from 120 yards to the hole, then I would go with the golfer that was better over the last 120 yards. I find it hard to rationalise my choice.