The logic of combining these three popular methods is clear, no discussion there. Still, they all work, or use as basis, the initial use case/ user group. With scaling comes diversity in user needs. Taking this into account early on allows much better informed decisions to create a more adaptive architecture from which you can develop versions that serve these different needs. The risk of all of the mentioned methods is that they create a path dependency based on the initial case after which small pivots are of course possible, but making fundamental changes is much harder/ costlier. Even better, by consciously anticipating on a wider diversity of needs you probably boost creativity for that architecture in the first place. Intentional Context Variation (i.e., “by design”) early on does not lead to more complex products , but the process has acknowledged that real life is complex.
I am really curious whether this line of reasoning is, or can be, recognised by fans of the lean and agile approach, and Design Thinkers. I imagine this recognition might be the case mostly for non-software based products but still, I wonder whether it can make sense for those as well.