It really does look like a hit piece. According to the article logic, we shouldn't trust TCP/IP because we already know it was a product of ARPA. The "trustworthiness" is misdirection when, really, we only need to scrutinize the code.