I found this to be a great preparatory course for "practical" topics in data science, such as linear models and machine learning. The homeworks and labs are well thought out, interesting, and very valuable for a good understanding of the material. There is also the bonus of getting to learn a bit of python (very popular programming language in data science) if you're not already familiar with the language.
To respond to older reviews of this course, the matters must have improved considerably with respect to submitting and grading the homeworks and labs. I didn't experienced any problems with my submissions, and I have completed successfully almost all problems (wrapping up the last two!).
With respect to the ratio of abstract theory vs concrete applications, I felt that the theory presented was necessary for a good understanding of the methods applied in the practical examples.
The following course was for me a great follow-up to Dr. Klein's course: HarvardX: PH525.2x Matrix Algebra and Linear Models (https://courses.edx.org/courses/HarvardX/PH525.2x/1T2015/)