It had its limitations and one of these in the matter of stress and strain was estimates.
Not really .. the numbers come out of equations although some of the set up conditions may require a bit of artistic licence at times. The bigger problem was the time and labour resource limitation. Hence one needed to be a tad more conservative to make up for not being able to look at a widget in sufficient detail.
With the advent of electonic computers which are great at doing LOTS of sums very quickly, the stressman can drill down a lot further into smaller and smaller parts of the structure. At the end of the process you can realise some weight savings.
Another consideration, if I recall correctly, is that some of the Airbus Types negotiated a reduction in the proof to ultimate factor which really gave a weight benefit.