This discussion always brings back a particular memory for me. When I was but a wee lad (about 4), I approached my father (a Duke grad from long ago) while he was doing yard work to ask him if we could have a pool built in our backyard. He put down his tools, walked inside, and proceeded to remove a few dozen documents from a filing cabinet including bank statements, tax documents, stock information, etc. Then he sat me down and asked me if I wanted to go to a good school, like Duke. When I said yes, he read through the documents with me and explained the importance of saving and how compound interest worked. Then we went through several pages of calculations on a legal pad to determine how much he'd have to save to send my brother and me to Duke.
Needless to say, I never asked for a pool again, nor much else. Dad did send my brother and me to Duke, though.
I think the problem will continue to get worse for about ten years, but beyond that time frame it's anybody's guess. The dynamics of higher education are going to change radically in the relatively near term. The early stages of the change are already visible. Many graduates find themselves saddled with debt and poor job prospects not so much because of the high cost of the education, but because of the disconnect between the cost and the proper incentives to lead them to select appropriate majors. Government-backed student loans have driven prices higher, but more importantly have created that disconnect and destroyed much of the market signaling that should exist in education. Meanwhile, cheap (or free), high-quality online courses are proliferating and gaining respect. They could radically alter the landscape of education for technology-related fields with solid pay. Many engineers and scientists still require physical tools that have usually only been available in expensive university (or industry) labs, and they will require university education a while longer. With the resurgence of "hacker" culture and local clubs with some of that equipment, I expect even that barrier will fall in time. At that point, the price of tuition will have to drop if universities expect to retain students, as the added opportunity costs on top of tuition will just be too high to ignore. In my opinion, expensive, private research institutions will face the largest challenges in this regard, so Duke might have something to be worried about after all, if not so much in the next couple of years.
How does all that affect planning for college? Beats me.