>>140
At the same time the Lisp master race churns out perfect code a couple of hundred times faster. But every evening they delete all the code they wrote during the day, because the world just isn't ready for it yet.
Well if you were as well read of PG as you say, you'd know that is almost the case. Viaweb made Yahoo Stores in Common Lisp with 2 guys in a couple of years, and then they sold to Yahoo. Turns out, they couldn't find CL programmers to work on it, so they decided to have a couple dozen C++ programmers reimplement it in twice the time. The rumor is that they actually resorted to implementing Lisp for some of it,
consciously confirming Greenspun's Tenth Rule.
The model of "high level stuff isn't used because programmers are uneducated about them and lack ambition." explains a lot more of what I see and personally experience than "high level stuff isn't used because it isn't actually valuable."
The problems with Lisp are cultural, both within and without, and solved in both cases by Clojure. Within - just needed a reboot. Without - just need a stepping stone, like Java. Java programmers are finally seeing the value of lexical scope and immutability, but their language treats this stuff as special cases. I think Clojure will look more and more attractive to Java programmers.