I disagree with Brooks, or rather he should speak for himself. What he says here may be true for the sizable branch of computer science that is software engineering, but most CS departments originally emerged out of mathematics departments, and more practitioners in academia deal in theory and experimentation than in building tools. The perception of computer science as an engineering field probably stems from the glut of undergraduates that go into industry as software engineers and thus represent the face of CS to most of society. As a result, computer science as a field tends to straddle both science and engineering. I don't think the author's attempt at redefinition here makes any sense.

It's certainly understandable that confusion surrounds the term 'computer science' though. It might be one of those cases where the name made sense when it was at first mostly a theoretical field, but maybe it's time for universities to start separating the engineering component into an actual engineering department. 'Computer Engineering' is already taken, though perhaps Software Engineering could be merged with it or placed into its own department. Of course, institutional inertia is a heck of a force so good luck with such reorganizations.

Is it really even useful to distinguish fields as engineering or science though? I generally follow literature in two fields: computer science and biology. In each field you'll find both methodology and domain application papers. The former contributes tools for study, the later contributes the result of study using tools. To put it another way, I know plenty of 'scientists' in biology, chemistry and physics who spend most of their time engineering instrumentation, and I know plenty of 'engineers' in computer science who spend most of their time conducting experiments and analyzing observations.

In conclusion, I don't find the distinction useful.