The difficulty is that years ago large corporations like ATT used to fund pure research facilities with no expectation that anything marketable would come out of them for many years, if ever. They were looking for Nobels, not money (in the short term). Bell Labs is an example. NV Philips still has a lab in Briarcliff Manor, New York. But years ago the accountants started demanding, well, accountability. And the scientists who were paid to sit around with their feet on their desks, creating new branches of mathematics and thinking their way to new knowledge, were asked to begin producing some product that could be sold.
I have friends who worked for Philips back in the golden days. These were the people whose random doodlings produced the compact disk. They were expected to spend much of their day staring out the windows and thinking beautiful thoughts about science. Today no one is going to finance that. There has to be direct market applicability in all the work the lab turns out. This means that only the federal government and sometimes major universities are going to generate novel concepts that are pure science with no short-term place in the market.
Yes, I have heard that argument frquently over the years. However, the federal funding agencies do not fund truly novel concepts. The people at NSF, for instance, will tell you that they do not fund new or "risky" research; they want to see published results first.
And to think Philips is hemorrhaging money now.
As long as industry is funded by wall street backed short term investors, not much money will be placed in long term R&D. It just doesn’t make good business sense.