The same could be said about most car manufacturers having done little to go beyond our current engines technologies.
So unless you consider Canon and Nikon to be research entities, I am not sure they can be blamed for the way they conduct their business.
Companies like IBM and Intel seems to focus a lot on basic research. I like to think that this is a conscious effort to be able to sell products that no other companies (or governments) have the capital and stamina to produce. This means that shareholders have to be in it for the long run.
A favourite story of mine on this subject was about Japanese vs Western manufacturers of electronic musical instruments. Yamaha wanted to get into the market, and from the early 70s onwards, they started introducing interesting, but commercially unsuccessful products. What they did was gather experience and build a name. At the same time, a number of exciting US and European manufacturers had creative and commercial success, but they had one problem: their investors were in it for the short profit. Every new product had to be a success, or else the company would go belly-up. Consequently, most of them (over time) disappeared, and those that survived tended to focus on safe, evolutionary products.
Then in the early 80s, Yamaha hit gold with the DX-7, based on research at Stanford. What it did was fundamentally different from what the competitors did, and it was a great success. Every 80s ballad has the DX-7 electric piano sound all over it.
Could this story happen today, with "landscape photography" as the product and Chinese/Indian manufacturers as the "underdog"? Does Western (and Japanese?) manufacturers walk the safe path of moderate evolution for each camera generation, while some revolutionary concept just awaits the right manufacturer to change the market? I think this is what the thread starter wants to discuss? My reply is perhaps, probably. But such advances tends to come from a radically unpredictable angle. Of course Canon and Nikon & Friends have a staff of PhDs that constantly think about S/N-ratios, Dynamic Range, MTF50, color accuracy and such. If there is a free lunch anywhere, chances are that they have simulated it, calculated the cost/risk etc. I dont think the potential from ditching Bayer is all that large. If we see such a product, I am guessing that (just like the DX7 in my story) will be more fundamentally different, and perhaps not "the ultimate high-quality". Perhaps it will allow flexible usage of multiple (cheaper) sensors/lenses, more room for dsp/raw development to play with stuff. Imagine being able to distort the sensor shape (or lense) to do local phocus/DOF adjustement .