Hey Tony,
First of all, I appreciate the transparency of recent partner program changes and new payout calculation metrics. I am a full-time data science, AI/ML technical writer on Medium and have been writing for three years.
I always prefer quality over quantity. Most of my articles are evergreen articles that have long-term value to readers. They get a very large number of external views from Google searches. Every external view is an opportunity to bring a new member to Medium.
Recently, I have created a course for "Neural Networks and Deep Learning" (https://rukshanpramoditha.medium.com/list/neural-networks-and-deep-learning-course-a2779b9c3f75), the first ever on Medium, as I think. It is a series of articles organized using the list feature which only includes my own articles.
Unfortunately, the number of READS (and also the Earnings) have significantly dropped since the implementation of new changes to the algorithm. It seems that the new calculation metrics are not good for technical writers.
As a technical writer, I often include code blocks, diagrams and prerequisite links of my own articles to enhance readers' understandability. I often include the prerequisite links at the beginning of my articles. So, readers click them and leave the page and then come again after some time to continue reading. In addition to that, the readers copy and paste the code blocks and test them. For that, they leave the page and come again after running the code. These types of activities cannot be tracked by the new algorithm.
In my opinion, member reading time is the most effective metric for technical writers. The referral program is the perfect solution for those who bring a large number of external views to the platform. It is also sunsetting.
These are my own ideas and suggestions. Thank you so much and I will continue writing high-quality articles for next years.