So much research in biology and chemistry happens in medical schools, where the online push is, at the very least, farther away. And much physics has to happen at large schools and national labs anyway, because of the tremendous resources it requires.

I wonder if there’s data on the amount of productivity (is publication the right metric, or maybe impact factor?) from different schools, so that we could get a sense of the size of the impact. Given that research productivity follows a power law distribution, and the best researchers are largely at the top schools which won’t be as impacted by the shift online, this may not be as large a problem for science as you imply. Although quite possibly a problem for all sorts of other reasons (universities provide a ton of benefits, aside from research, for the communities in which they’re located that could be lost).

One last thought – you move to government as a solution, but there’s a growing movement for crowdfunding (kickstarter for research) that could be an alternative. I agree that science is a public good, but I worry about putting it in the hands of politicians more than it already is. And as Wendy notes above, we should be wary about spending taxpayer dollars if other alternatives exist.