Ingura,
You are so correct! I have been building a model working from blueprints from the US National Archives. They are on microfilm, and the cameras used to photograph the original blueprints had some spherical distortion in the lenses so the microfilm images are slightly "pincushioned." In addition, they couldn't get an entire blueprint sheet (up to 12 feet/4 meters long) in a single photo, so there may be six overlapping photos of a single drawing.
To make matters worse, the microfilm digitizer I use also has lens distortion, and I have to copy each microfilm image in four to six scanned images in order to get resolution good enough to read the fine print. I then piece the images together in Photoshop, up to 24 scans to make a single drawing. I use the Warp and Scale functions in Photoshop to correct pincushioning and fit all the parts together. The result is a very good but not perfect reproduction of the original blueprint in high resolution, but in many cases there is obvious distortion.
I have done this for two ships, a Cleveland class cruiser and a Baltimore class cruiser. In both cases the original blueprint hull line drawings were obviously in error. For both hulls the fore and aft station lines drawings disagreed by more than two feet for the width of the hull! This resulted in a one foot "kink" or stairstep in the hull plating amidships - obviously wrong! I discarded the hull lines drawings and worked directly from the Tables of Offsets and Shell Sight Edge tables and reconstructed the hull lines without the errors. The real ships were built from the tables, and not from the drawings.
Finally, the original blueprints were not necessarily drawn to scale - in fact, the general deck plans and profile views often are just sketches, and the dimensions shown are different from the dimensions given on the individual detail blueprints. Add to this the photocopier distortion you mention, and the drawings are of questionable accuracy. For this reason I do not use the lines in the drawings for modeling. I always try to use the dimensions given in the blueprints (or measured on the actual museum ships) and reconstruct the parts to scale.
The errors in the drawings don't matter at 1:700 or even 1:350 scale - they are less than the accuracy you can achieve modelling by hand. But at 1:100 or larger the errors can accumulate to produce noticeable imperfections in the model. At large scales it pays to work by the numbers, just like the real shipbuilders did.
Phil
http://www.okieboat.com
A collision at sea will ruin your entire day. Aristotle