An interesting article here that claims that claims that KM delivers only efficiency improvements, and not quality improvements. This is based on a study if several IT projects, looking at their use of a "KM system". I quote
The research results show that mean team knowledge use is positively related to improved project performance, but not ‘significantly related to the quality outcome’. As the paper says, delivering defect-free code requires that team members solve their assigned tasks, but ‘also successfully integrate the tasks’.
However whan you look at that they mean by a KM system, they mean a database containing documents, white papers, and bits of code. This is a long long way from a complete KM framework.
In a way it is good news that even this truncated and stunted KM system delivers some benefit. I wonder what they would have gained if they had used a full KM framework? I bet they would have seen quality rise, as well as efficiency.