You make an excellent point about social convention, and I completely agree. However, one important thing to note is that SW is not involved with core Mathematica development, he does research and remotely manages his company. He suggests high level features that are then implemented by the kernel team (which I was on). One incredibly annoying aspect of this bureaucratic development process is that you never know when your kernel project will get killed by someone higher up the food chain. Obviously, this happens at many companies, but when it happens as often as it did when I was there, you see key employees leaving a lot more often than you do elsewhere.

It's worth pointing out that the architecture of Mathematica was designed by Stephen, and in my experience with other large software projects, it is by far the most elegant and well-designed. Reading the original code base gives every developer a great appreciation for the genius of Stephen Wolfram, who is a completely self-taught computer scientist. So he definitely deserves credit for what the language is today, because if he hadn't planned that far ahead and realized the need for a general purpose computing platform, the pace at which the Mathematica team can add new features would not be nearly as fast.