There are a tonne of these post doctor research guys that study these kinds of things all the time at universities all around the world. At my uni too (in Oslo) we have quite a few really interesting doctoral disputas where they have done a lot of research into subjects that could really help game tech leap forward.
The problem, is that what you see in unis is usually far away from being practically applied (thus the term academic, look it up), and you have come to face the fact that there is almost always some side effect that would typically keep the tech from reaching the market. One such thing is technology, of course, as a lot of unis have special hardware and software to play with, along with tonnes of students that would gladly help out with research projects. Another thing, believe it or not, is patents.
More and more unis are understanding that patents are really important, and being research organisations, they file patents all the time. Got a good idea for a PhD software framework? File a patent. A clever master student wrote some great, albeit not ground breaking code? File a patent. One of the professor IIs we have at my uni mostly works at Bell Labs in the UK, and I recall he said that during the course of a year it is common for a researcher to file between 3 to 5 patents. So if it really comes to the point where there is a great idea that can also be applied in practice, then on top of the practical problems there will at least be a couple of years before the idea can be used freely, because of the patents.