It's an unfortunate self-fulfilling prophecy of the modern data-driven mindset. Sometimes the very act of collecting and acting on that data materially affects the data itself by creating perverse incentives.
The two parent posters are far from alone in my experience. Lots of people are getting turned off by the variable quality and uncertain future of the home-made productions, which means lots of people are holding off starting to watch a show until it's somewhat established and had some positive reviews. If your management strategy is to measure early engagement with your own shows and viciously kill off anything that doesn't make the cut, and if your viewers know this, then you have defeated yourself no matter how good the show is or how popular it would naturally have become.
A few years ago there was almost a trend for shows that weren't getting the numbers to get wrapped up with some sort of mini-series or TV movie so at least there was a chance for the production team to finish telling their story and give some closure for the fans (who might be fiercely loyal in sentiment even if too few in number to sustain the show). It's a little ironic Netflix would probably have been in a better position than anyone to adopt this kind of strategy and establish a reputation for being trustworthy and loyal to fans. Now it has the opposite reputation and we're openly speculating about whether it will ever recover.