The most common way of measuring project success is by relating to the ultimate value of the deliverable. I think comparing achieved performance provides useful feedback for improving the accuracy of future project plans.
At the same time, we all know that projects are all different, so a direct comparison is always risky. What is your opinion?