Theory: Broder didn't realize the logging capabilities of the car, and when the Model S' software ui initially supported his internal baises he took liberties with the truth. By "documenting" his experience through Tesla support he attempted to falsely add credence to what would be a traffic generating, "anti-electric" review masked in the journalistic repute of the NYT.
Firstly, all of Broder's excessive winging about the cold weather (I think) was designed to subtly imply that the Model S doesn't work in the cold. You future buyer, will be cold and your car will break. This is why Musk had to address the cold weather link directly in the evidence blog posting.
Secondly. Broder likely couldn't have fathomed that every parameter in the car was being logged. Very specific details add credibility and character to a story. They make the author appear diligent, and one who gives great attention to detail. In the past such details were a "literary tool used to bend the story. Now thanks to data driven engineering words and truth in such matters should align more closely in cases such as this.
Lastly. For a man who may or may not have a bias against electric vehicles (cars at least), the observation that "the estimated range was falling faster than miles were accumulating" at the outset of the author's journey might have set the tone of the coming review. With all the incessant calls to Tesla support to document all the "trouble", Broder had plenty of documentation to support his (what was IMHO a) journalistic malignment. This angle also had the added benefit of generating views for NYT - plus through the courtesy of Tesla arranging a tow - the money shot.
I hope NYT has the ethical chops to do what they must.