Spinoffs have a long and sorry history in NASA advocacy. The problem is that advocates went from "NASA had some minor role in development of technology X" to "NASA is entirely responsible for technology X". One sees this over and over again, to the point of the claims being outright lies. "NASA gave us integrated circuits" is a great example.

Claims that space spending had a 8:1 (or 24:1, or whatever) payback via spinoffs have typically just assumed the spending has the same macroeconomic payoff of private R&D spending, which demonstrates absolutely nothing.

Without that contribution inflation, one has to somehow show that a particular technology would not exist without the putative spinoff contribution. And that's really difficult to do in most cases. Here, the military has had an interest in high power millimeter wave sources for a variety of reasons.

The argument remains, also, that if spinoffs are an inherent part of research, then it doesn't matter what the research is on, and so you get the highest payoff by funding research that makes sense via direct benefits, as the spinoffs come regardless. You'd need to argue that fusion is somehow better at producing spinoffs than other research. I find that likely; I suspect small scale research is more likely to provide spinoffs, as desktop technologies seem likely to have more applications.