The problem is not so much the total time spent on the device but more the time spent per content nowadays. There is a certain value in spending 3 days to accomplish something whereas imo no value in spending 5 seconds per content.
This guy probably lives in his own small world. If you want to publish in PLOS as a researcher from say Turkey or Uzbekistan or any other country where the value of your money is nil, you might easily have to pay your yearly salary or half of your funding to get a single paper published.
I am always surprised how my first guess gets wrecked by Bayes rule. I would have thought that there is 5/6 chance I am on side track and 1/6 that I am on the main track.