Indeed I have noticed textbooks often only contain minor changes from version to version. In other cases, entire subject areas can change as research progresses. But sometimes the only way you get to know that is by looking at older texts.
The quantity of changes may be less relevant than the quality of the changes: i.e. did anything major change in this field in the last n number of years. With a closed source software program, like Windows, it's very difficult to assess the quality of the changes. All we see is the user interface.
There's a difference between a list of "new features" (staying with the Microsoft example) and being able to diff the source code against a previous version and evaulate the changes for yourself.
It's like the syllabus that says you need the latest version of the textbook but does not tell you why the previous version will not do.
Look at it this way: in "most cases" (whether it's software or textbooks), from version to version, there's more substance that stays the same than substance which changes.
Not sure about you but I'm a little wary of buying the latest version of an expensive textbook in order to gain a small number of reatively minor changes that I could easily identify, and take note of, by comparing the latest version with the previous one.