Probably can't happen without huge changes to the tax code IMO.
That said, I think bringing the amount of research science we have under the umbrella of academia has been bad because it's basically introduced a plausible deniability and reputation laundering layer that furthers the 3-way revolving door between academia, government and industry.
IIUC, industry does get R&D tax credits- but that's probably not a good incentive system for basic research in industry.
What's happening now is that a small number of science-friendly rich people are making foundations/institutes that carry out basic research (Zuckerberg/Chan, Schmidt, and a few others) but it's unlikely those will completely supplant academic research at universities funded by NIH/NSF/DOE.