I have been following Judea Pearl's work and I am excited to finally see him publish a book on this topic. I have been observing these issues at the lower end of machine learning for years now and it is amazing to see someone at the very top confirm them in this book.
This summer, I will be starting a research reading group as well as a corporate entity to help change this outdated status quo. If you want to be involved, let me know. The work will focus on three frontiers: public research, pro bono advice and private consulting.