I remember a friend of mine who plays poker telling me that he didn't do that well at a particular version of poker called Omaha until he started looking at other people's mistakes and making sure he avoided making those mistakes too. After that he started consistently winning. It's certainly often as useful to look at things that don't work as to look for some secret formula as to what does work.
Recently I was describing to a colleague an experience I had trying to learn some computing stuff - basically I knew too much of it already which made the whole experience a bit boring. I loved the way that she put it - that it puts you into a 'skippy mood'.
Once you've got into the habit of skipping over large sections of a book, it becomes hard to go back to reading it properly. This is something that can be a teensy bit hard to avoid - I guess you need to be fairly specific in your readership for a start and not trying to be all things to all people. A book on web services using Java for developers doesn't need to start with a ten page description of what the internet is and why the internet is a wonderful thing for instance (unlike the one that's sitting downstairs that I haven't quite got into yet) .
There's also the art to letting people skip sections without getting into a skippy mood. I think if you can work out the right place to start without having to really read any of the text then it's much easier. This is where good headings and so on that can be scanned easily can be really useful, or just plain out explicitly telling people it's ok for them to move onto the next section if they already know certain things. I wish people would do that way more.
The other thing I've noticed myself sometimes doing is following instructions but not really taking in whatever it is that I'm supposed to be taking in - essentially going onto autopilot. Science practicals at school were like this for me, maybe because we never really designed the experiments ourselves. I became an expert at getting an A grade for practical write-ups while most of the scientific principles washed totally over me.
Interactive buttons in museums are another classic example, and I've managed to enjoyably play board games set in all sort of historical periods without taking in anything at all of the historical setting.
Perhaps sometimes though you've just got to hope that maybe it won't help everybody but it will help some people. I know I learn more by typing in somebody's code than just reading it and I'm sure people have learned stuff from computer games rather than just the right sequence of buttons to press. But I think it's easy to fall in the trap of assuming that just because somebody is doing something that they must be learning everything associated with the context.