I've written code that shipped in 1 to 2 million units. I also did an academic post-doc in computer science at a good university.
Testing is not just underappreciated in the academic community it is widely misunderstood. Academicians are fascinated by questions like how do you create a program to test your programs. By comparison the automation of human designed tests is not academically worthy. But this is most of what separates commercial grade code from research code. It is what creates 99% of the economic value of testing. And done well it elevates the status of testing from clerical work to top-tier design work. This book explains why this is true. Armed with that insight, a skilled technologist will be able to work through the rest of the details.
Nevertheless, the book offers quite a few recipes. None of which can be implemented without a skilled technologist capable of working without a recipe, but all of which enhance the value of the book.