Our different perspectives about the nature of DNA do indeed account for some of our disagreement. I don’t consider DNA to be a hodgepodge of genes in no particular order interspersed with vast amounts of gibberish. Instead I think that the majority of DNA appears to gibberish because the current scientific understanding of DNA has barely scratched the surface of what is really there. It only makes sense that scientists would figure out what the protein-coding sequences of DNA do first. It doesn’t make sense to jump to the conclusion that the rest of it is junk.Randy has taken the safe approach by claiming that the DNA that is currently considered "junk" by geneticists will eventually be found to be important. It's safe because it's a position that can't be falsified. There's no way I can possibly prove that "junk DNA" will never be found to have an important function. He can lurk in that nebulous "one day we may discover its importance" realm forever, because he has until the end of the universe to be disproven. All I can say is that since he can not show that "junk DNA" has a function, there's no evidence of a function. Generally speaking, the burden of proof is on the person who asserts the existence of a thing, not the person who denies it. To use a cliche, I can assert that there's an invisible, intangible dragon living in the shed behind my house, but do you have any reason to believe my assertion? Without any evidence that junk DNA serves a purpose (beyond adding to the medium in which mutation and subsequent selection can work), there's no reason to assume we will eventually find one. So "Yes", Randy, it does make sense to assume that DNA that has no known function is junk.
But let’s say for the sake of argument that you are right about the nature of DNA. Even in that case my argument is still valid. As you recall my calculation regarding the probability of information used a 151 character sequence. The average length of a gene is 1210 base pairs and is hardly equivalent to the three of four letter words used in your example.
As to the rest of his case, no I don't see the validity. The average length of a gene may be 1210 base pairs (I haven't done a fact check on that), but the current average length of a gene is irrelevant. There are certainly many shorter genes, and there's no reason to assume that the current average must apply all the way back to the original molecule that started the process of evolution. I've already linked to a good explanation of abiogenesis, which works with self-replicating molecules that are composed entirely of junk DNA. You can get the ball rolling with pure static, so you don't need to have hundreds of base pairs in some kind of functional code for evolution to begin. From abiogenesis forward, any DNA strand that forms the simplest of productive genes is gravy.