@User: My original pessimistic assessment was based on the fact that the single board sitting on my desk was not one of those 40% of chips that work. Based on my tentative conclusion that there might be enough variation over the wafer that some chips would function, we mailed out all the boards. Fortunately my intuition was correct!