The phrase that leaps out is
If cases need to be bounded by the researcher
Why do they? Bruno reckons that you let the actors tell you when to stop.
Perhaps the best way in is to ask the question: What’s going on here? Now the word here does suggest some sort of limiting of the researcher’s gaze but what she/he is really after are the practices which enact the various realities of the situation. So it’s not the case that the researcher with the god’s eye says, “I need to work out what’s going on here.”, but rather, what is being done in the performative sense? To take examples from Law’s work: how are salmon being done? how is mad cow disease being done? how is the train wreck being done? etc.
I think that it is useful to skip ahead a bit and read the paper he did on syncretism which I think I, or others have mentioned previously:
Law, J., Afdal, G., Asdal, K., Lin, W.-y., Moser, I., & Singleton, V. (2014). Modes of syncretism: Notes on noncoherence. Common Knowledge, 20(1), 172-192. Retrieved from https://www.sv.uio.no/tik/personer/vit/kasdal/15_ck201_modes_of_syncreticism-libre.pdf
Looking for boundaries is, to me, an instance of what they call a bias for purity, for neatness, tidiness. The job is not to tidy things by slapping boundaries around stuff but to ask better questions, like given all these different realities being enacted, how does it hang together, if in effect it does? When stuff appears coherent, unitary, smooth, work is being done (practices) so that it appears that way. That’s really interesting. What is dull are accounts that represent a single reality that coheres and makes sense. Hence the title, I think, making a mess with method.
And if you want a paper that does a nice job gesturing to the infinite regress that I suspect is behind your question, then his paper on collateral realities is a gem. Also not bad model of how to write a paper by attending a conference presentation
Law, J. (2012). Collateral Realities. In F. D. Rubio & P. Baert (Eds.), [i]The Politics of Knowledge[/] (pp. 156-178). London: Routledge.
Hope that helps. Not sure I am connecting or just shooting the breeze