I disagree completely with the author's point that the ad model is not regressive. The author points out correctly that charging everyone some $ is regressive, because for some people that amount of money is a lot, and for others that is a little. Totally makes sense.
Then, we go on to ads. Ads charge everyone a similar amount of bandwidth, attention, etc. But you know what? Some people who have more resources or knowhow will understand how to block ads. And they probably won't be paying by the MB on a crappy cell phone plan such that they spend their money on bandwidth to load the ads, while the content they want to read languishes below the fold of the ads and they struggle to navigate to it on their crappy device struggling to render ads.
The costs are more abstract than when paying in actual dollars, but surely we can recognize that the cost of ad supported web pages is also not felt evenly by everyone. As a privileged software engineer, I can guarantee you that the impact of "paying" for things with ads is felt far less by me than many others. That is regressive in my opinion.