Nick Matzke's long comment in the preceding thread inspires this post. Redundant Genetic Instructions in 'Junk DNA' Support Healthy Development is a Science Daily article. From the link:
ScienceDaily (2010-07-17) — Seemingly redundant portions of the fruit fly genome may not be so redundant after all. Repeated instructional regions in the flies' DNA may contribute to normal development under less-than-ideal growth conditions by making sure that genes are turned on and off at the appropriate times, according to new research. If similar regions are found in humans, they may hold important clues to understanding developmental disorders.
Things may not be what they at first appear to be. Seemingly functionless repetitive sequences may regulate regions with already identifiable roles. The size of non-protein coding regions was explained by the activity of transposons alluded to by Nick. If the size of that region in turn is eventually found to be more functional then initially assumed we will have witnessed another surprise and a theoretical adjustment already evidenced. "Junk" would serve as a reservoir from which future function is derived. As the SD author put it, "scientists increasingly believe "junk DNA" is crucial for turning the information encoded in genes into useful products." Of course those information rich genes already were functional right?