Hi,
I totally get your points. All of the numbers reported in the tweet are total compensation (TC). Obviously there are some problems with TC in that it depends on stock appreciation, and how people report it, but that is the best I could do with the data available.
I also tried to address the different levels by including coefficients for levels in the regression. So if you look at the table at the end, you can see senior engineers made more than SWE 2s. I thought of including interaction terms, but I don't think I had enough data to get reasonable estimates for that, and it would be harder to interpret.
That's part of why I was looking into this, is I wanted to understand how things might change in the future. I have no special insight on what will happen, but there still seems to be a lack of Senior Data Scientists. Of course, that might change in a few years as people get more experience.