"Isn't the entire rationale for providing government funding for academic research premised on the idea that scientific knowledge is a public good, to be shared

"Isn't the entire rationale for providing government funding for academic research premised on the idea that scientific knowledge is a public good, to be shared freely?" No it isn't, it's only a part of it, and only applicable in some cases. Other rationales include stimulating research on things that are considered too risky by the mainstream manufacturers, so as to facilitate radical innovations (in the Geoffrey Moore meaning of that concept). Under those circumstances, allowing patents on the results is required otherwise the people doing the research wouldn't have a reason to do it in the first place. Plus, the grant money is only a part of the eventual results; the original ideas, the extra labor, outside financing, etc are equal contributors.