Hi,
In my opinion, the only useful thing about the blueprints are the techniques of model building. If you study the model, sometimes you learn a new feature or combination of functions.
I don't think you could use them for their un-real assumptions. Last blueprint, I was looking at, was "financinal services branch performance" which - according to cognos - could be used for any retail bank branch profitability. The bank in the model had a product calculation with 20 lines, has only 7 products in total!, 3 segments, and 20 branches (which were in e-list so could be increased). If I take this model and use it in our bank, it will never work, because, the calculation has 10-120 lines, 150 products, 400+ branches plus hierarchies. They make all the cubes really simply with all things in one place, but if they will increase the sizes to real measures, it will never work.
I've seen something similar with other models, where it's nice to get throught them to get a different insight on things you know, but than you only use simple elements of it.