I asked the folks at Carl's what the target stretch should be for my 126 inch material. This has always been a mystery with answers that are not numerical. Finally the answer was given 129 inches stretched, I don't think that's an engineering answer more experience.
Applying that in general yields a stretch of about 2.5%. So that makes the ideal perimeter for my prototype 78x44 material 80x45 stretched. This is significantly more than I was stretching,
Applying this much stretch yielded an acceptable result, The result is still not perfect because I did not have a way to measure horizontal stretch while stretching vertically. It is easy to measure the vertical stretch because I have a 45 inch vertical target line. However, stretch is being applied along the line at the same time with no way to measure the stretch along the line.
For my real screen I will mark a 129x72.5 rectangle on the frame as the target outline of the stretched material. Then I will divide the material into eight segments and mark it. I will divide the larger target outline into eight segments and mark it as well. Now I can line up the marks as I go to make sure the stretch in both directions is consistent.
This might sound like a lot of work but I expect that it will be half hour well spent. I get the sense that if I do this enough times I wouldn't need the marks but since its a one time thing this is the only way to get it right.