They're not mutually exclusive. For example, I've done consulting work for a private geothermal energy LLC (AltaRock) that does a fair amount of R&D funded by the DOE (and maybe NSF?) and other public sources, though I believe that the company is also venture-backed.

A lot of the hard part of the development in geothermal is fighting with problems that aren't very solved scientifically, particularly understanding the dynamics of stress and temperature in the upper several km of the earth's crust near active volcanoes. I was working with AltaRock on these issues, which are chewy enough that NSF doesn't consider the baseline levels of knowledge good enough to fund research in, as NSF wants science to be at a point where some sort of obvious scientific outcome is highly probable before they will fund it.

I assume that this is pretty widespread in cutting-edge R&D: Those involved find themselves working at the ragged edge of the known universe, even if both academic scientists and the general public think of the topic as pretty mundane.

So I view a fund like this one as quite complimentary to public funding: At some level, venture-type funding that has a high risk tolerance can help get projects off the ground to the point where they can get funding from both more conservative government funds and more typical institutional investors.

(Full disclosure: I also own a private geoscience research company that solicits a lot of public funding as well as doing consulting and non-fundable research, but it's really just a shell for me to try to get paid to do cool science, rather than a nascent company focused on growth.)

The article directly addresses your question:

  >One thesis driving the fund is that only governments have the resources 
  to invest in fundamental [clean energy] research, through government labs and 
  funding for university research, at a scale that is needed for breakthrough advances.
  Government funding in areas such as battery technology and materials used 
  for solar power, for example, has seeded the growth of those areas, with 
  their commercialization subsequently funded by private investors.

Let's restrict ourselves to ultra-long term funding, i.e. that with 20+ years horizons.

In this view, there are more dollars in the public coffers. But they are more politically entangled. If your project is a political darling, e.g. a clean-energy company with a need for lots of factories in lots of districts and a cap table of retired Senators, this is close to frictionless capital.

If you live in reality, however, there are costs and political risk attached to this capital. The private pot is smaller, but faster-moving and less unpredictably politicized.

A balance between the two would be good, if only to encourage the public coffers to become more efficient.

NSF budget is almost 8b per year. This fund is 50m per year. Its simply not on the same scale. Science fails without government funding.

While we're at it,

> which Gates and his fellow investors describe as a market failure that large-scale, long-term private investments can address.