I agree with the underlying logic of your argument, but see the percentages as quite arbitrary. What would you consider the margin of error (perhaps "error" is the wrong word here) for your percentage estimates? Surely you are not professing to be certain on the exact numbers.