"Which is precisely my point. The author has selected a particular sequence and shown that that particular sequence is very unlikely to occur, just as I did. It therefore doesn't support his conclusion." - edsheppa
It might be your point, but that isn't an accurate representation of what Watson did.
Watson demonstrated the math for calculating the probability / improbability of data sequencing itself into a desired output (e.g., Life, the first sentence in Shakespeare's Hamlet, et al). What you did was demonstrate that some output will always be possible.
Sure, you flipped a coin a thousand times and got an output result. That isn't exactly unsurprising. Anyone who flips coins a thousand times is going to get an output result. The odds are 1 in 1 that you will get an output, not 1 in 10^300 as you wish to imply.
On the other hand, flipping a coin a thousand times and hitting a pre-determined output is another beast altogether.
The difference is about the same as seen in a powerball lottery. The odds are 1 in 1 that every powerball lottery machine is going to select an output. Hitting that output twice, or predicting that output in advance, however, brings up entirely different and much lower odds, obviously.
It is most assuredly an accurate representation of what he intended to convey as evidenced by his conclusion. I think you agreed to that a few posts back.