[delian66]

I disagree. The observation that working complex systems evolved from simpler ones is very useful, because from it follows a rule that, if you want to make a complex system, your best chance is to make a simpler system, then iteratively improve it. It practically solves the analysis-paralysis problem that otherwise is inevitable.

Even total redesigns will benefit from the experience and information, that you got from the simpler system(s). These are often hard to derive by other means.