Jones is a master at data collection, distilling it, and drawing supportable conclusions. Like his other books (especially Estimating Software Costs, ISBN 0079130941), this one is wide in scope and deep with data and techniques.
He begins with background material on software process assessments, comparing his company's technique to SEI's, and correlating the two. Note that Jones' approach predates the one developed by SEI and was first published in his 1986 book "Programming Productivity", ISBN 0070328110. This book is a natural extension of that earlier work.
The next part of this book is an exhaustive survey of benchmarks and baselines, including pitfalls and an interesting discussion on activity-based software benchmark data. This material is a lead-in to 36 key factors that Jones identifies, including software classification, project-specific, technology, sociological, ergonomic, and international factors.
Subsequent chapters address best and worst practices, process improvement, and benchmarks and best practices for various software classes and development approaches, including internal IS, outsourced development, systems, commercial, military and end-user software development and delivery. Each class is treated in a comprehensive manner and the findings are well supported.
This book is an ideal resource for any organization wishing to establish a baseline before implementing initiatives such as CMMI, SPICE, etc. More importantly, much of this book is as applicable to the SEI assessment approach as it is to Jones's SPR methodology. I also recommend using Software Program Managers Network (ASIN B0001M00RA)in conjunction with this book (paste the ASIN number in the search box at the top of this page to reach it).