Option 1 would have some very interesting consequences. Seems like there'd still be a mandate to publish NIH funded work as open access, but no funding to support it.