Gilb introduces three ideas that may not be new, but extremely useful for software projects. Evolutionary delivery, or delivering small parts all the time, while getting feedback; formal inspections, which will save time and hassle; attribute specifications, to help you set a clear goal. I guarantee that this book will see software in a fresh, new perspective!
Gilb covers only a small portion of "Software Engineering Management" in detail. Other important issues are either not addressed at all or receive only cursory treatment. Still, what it does go into depth about it usually well thought out and makes imminent sense. I'd say it's worth reading, but it may be a bit expensive for it's narrowness.
This book covers three main topics: evolutionary development, requirements management, and inspections. A good introduction to all three, but if you plan on trying evolutionary development, I also suggest reading the pertinent sections of Rapid Development by Steve McConnell.
Gilb's principles of software engineering management is a forerunner and influencial book to the agile software development revolution. When you read this you see all the parts that you recognise from elsewhere, plus learn how to measure your success using impact estimation tables. This book is all about incremental iteration and, even if you've read his newer comparative engineering book, you'll still find lots of useful details here, which are not in the other book. This book is delivered in a a more relaxed manner, and also covers a few useful aspects such as interfaces, and antecedents of the agile perspective in the literature, which is rather nice. Definately one you should pick up and read.