While you're at it, I'm also curious how you think cutting profitable departments (such as CPFB) or white-collar law enforcement helps anyone other than the president's friends.
No, public research is an investment. The job of the government is to spend money on things that help the country's people. That includes roads, the military, and science.
Even then ROI for taxpayers is staggeringly huge in the long run. Calling it a "net-loss" is ... well. It is flat-earth level thinking.
The OP asked for an answer, and the fallacy I’m trying to point out is this assumption that “all” research is generally worthwhile. And I don’t mean just that there are exceptions, but swaths of topics and institutions and researchers and methods that aren’t really adding anything of value. Some areas, like aerospace, have been the focus of intense research for a long long time. This is to the point where SpaceX has been able to walk in and build a space program from scratch. I don’t mean to minimize their hard work, but it just shows that we’re not in the 1930s anymore. It makes sense that you should occasionally adjust your priorities from time to time based on your needs.
Private investment is risk-averse: they will support reliable short term gains over dubious long term breakthroughs. That's fine, but history has shown it's not the best way.
I don't think all research is worthwhile; I think grant proposals should get shot down if the lab/idea isn't up to snuff.
I'm happy for companies to do R&D! There's lot of examples over the years (Bell labs, Google, etc). I just don't think it makes sense in high risk situations with long reward lag times. The exceptions will be the really, really big companies flush with cash, which has its own issues.