I had gone into the DPhil with the idea that I was going to create a framework for RSE that distinguishes it from "regular" SE. What I had to come to terms with is that I don't know (indeed, "we" don't know) enough about RSE to build that framework. The advantage of the grounded theory approach I have settled on is that it allows us to build an explicatory theory based on what practitioners and stakeholders themselves think of RSE. So I will be able to explain how research software engineering participants think about RSE, which will be a useful basis for constructing or evaluating any later framework or other intervention.
The lesson here is when you come to a problem domain with a proposed solution, particularly one that's based on your own experience, you need to be able to show the justifying argument for adopting that solution. If you can't, that's not a problem, it's an opportunity to do the research: it's an open question.