I studied college level EE for five years, but that was 30 years ago, and I've been doing mostly micro-electro-mechanical systems (MEMS) for the last 20 years.
Like General_re, I've forgotten most of what I knew.
The core argument isn't a quibble about diodes and transisitors, it's whether a useful circuit can be designed by an evolutionary process.
The answer to that question is yes.
Well, it may be to you, but here we are discussing a specific circuit. Since you have a background in the subject, you might have some opinion as to the evolved circuit and how it functions. Specifically, how it functions compared to the patented circuit in the environment and uses for which it was designed.
Computers are very good at iteration given a problem, a desired solution, a dbase and a good programmer.