Maybe I'm getting a bit too detailed, but did you find that the variants of programs used by the end-user sometimes varied so much that it was hard to really apply to that testing?
I was part of an organization that looked at digital literacy, for example. We found that it was really hard, because there are so many different types of programs. There are legacy programs that people have been using for 15 years, because they work so well. Sometimes, when different scripts change, it becomes more difficult.
Was there a way to get over that challenge?