Yeah, what Adam Ierymenko said:-) about hitting a complexity limit being not at all synonymous with stopping progress. Except that I was going to say "computer programmers" instead of "engineers", and I was going to use the example that when duplicate functionality in the mitochondrial genome and main-cell genome gets replaced by shared functionality, the organism tends to win back some ground from the the Williams limit you described. And, incidentally, the mitochondrial example is very closely analogous to something that practicing computer programmers pay a lot of attention to: Google for "once and only once" or "OAOO" to see endless discussion.