One problem is that the industry today is ruled by Wall Street and has very short term outlook. We know for a fact that most industrial giants have closed their research labs or shrunk them greatly. Just for kicks, which industry will subcontract a CERN collider or a Hubble telescope? We are also seeing this in biomed. Industrial firms were in no rush to develop Ebola cures because they could not see the profit. Now the government is giving tons of money to the few promising leads trying to play catch up and we are losing lives in the process. Similarly, and more ominously, companies are not investing in new antibiotics and we are seeing major antibiotic resistant strains arise and threaten medieval-style misery. I suppose we will dump money at the problem when the first few millions die from some new pandemic caused by a relative of a previously benign bug. In short, bringing in industry to manage science is a terrible idea and there is tons of examples right here right now.
National labs could certainly be diversified in their mission to facilitate transition of academic minds into industry. The problem is... where are those industry jobs?
Putting everyone on GSA scale is a great and overdue idea. You will have to boost grant funding to prevent existing research projects from grinding to a halt but after that boost you could maintain that level steady. NIH already has some salary guidelines but they do need to be boosted.
In any event, the real problem is the lack of funding and hence jobs, whether in the industry or in academia. Personally, I feel that the solution is to acknowledge that we have too many graduates at every level and to then dramatically increase academic standards so that only very few could get a PhD and this degree would be seen (as it once was) as a major accomplishment that truly sets one apart from their peers. I think that if we simply produced ten times less PhDs then we would have none of the issues with postdoc glut. In hard sciences, we should eliminate Master's degrees because right now you get it if you are a failure and cannot get a PhD. The degree might have its uses in Engineering though. We need to cut the number of bachelor degrees until salaries for tech work start to go up. We also need to reorient scientific labs to employ technicians, rather than students or postdocs. That way labor will get a market price and will not be tied to a degree. Bachelor students will once again know that there is an industry job waiting if they can get a degree. PhD students will then be apprentices who will know up front that they aim for academic freedom but their odds are like the odds of winning the Olympics. And if PhD students fail to get a degree then they will still have a bachelor to fall back on.
Most importantly, we need to dramatically increase salaries for middle and high school teachers of science and math and simultaneously increase hiring prerequisites. This is where we need to channel the current glut. The brightest people on the sidelines need to be channeled into getting our society as a whole up to speed.