Things are not as simple as stated in the book: the nebula expands rapidly at first, but then at a slower pace. Thus, you can't just divide radius by velocity as stated in the book. (I think they just wanted to give an insight how that can be approximately calculated. People often think that scientists and astronomers get some information just using some "special instruments" (time machine for age of nebulae, for example) and the author tells us that this can be achieved pretty logically.)

The average velocity is thus bigger, and total time thus smaller. If we say, for example, that the average velocity is around $100\frac{km}{s}$, then we get $9000\text{ yr}$ for largest nebulae with $3\text{ ly}$ of size (but smaller for smaller nebulae), which agrees with the book.