Sakasuka,
I absolutely agree with you about the fact that there are a huge number of factors that effect clinical outcome. However, i think that for diagnostic purposes the "atypical" classification is very useful. Doctors are faced with the perhaps daunting task of diagnosing someone who presents with certain symptoms. For the health of their patients, they need to be able to differentiate clinically between different disorders. CF symptoms can be extremely similar to multiple other disorders. The atypical CF classification helps doctors differentiate these symptoms to either make a CF diagnosis or not. There are a few similarities among people who are classified as atypical CF: they tend to have lower sweat chloride levels, may present at an older age, may have some mutations considered "mild" and tend to have symptoms in just 1 organ. Doctors have a difficult job and they are not always that great at it. They depend on diagnostic algorithms to figure out what is going on. The atypical cf classification is really about that and NOT about predicting absolute clinical outcome. It is so overwhelmed doctors or doctors lacking real knowledge/experience of/with CF will have an algorithm to follow when making a diagnosis. It is about procedure.
When it comes to who is doing better at better down the line, you are right, that will depend on environmental factors, age of diagnosis, patient complience, modifier genes, nutrition and excersize.... a TON of things are involved.
ps USUALLY all siblings (unless having different moms or dads) will have the same mutations (ie but not same modifier genes necessarily). The fraternal twin example mainly controls for environment....(since they are not identical) and even then there can be difference in patient complience at some point.