Graham and his cofounder (and later Trevor Blackwell) all wrote different areas of the product, in different languages.
Joe Armstrong (creator of Joe's Abstract Machine) has also argued very recently for this kind of development.
I'll be the first to admit that I haven't seen it work myself. But when the people practicing it are at that level, maybe the problem is just that there aren't enough really good programmers?
I wish the world were full of really good programmers. But when you don't, you have to make do.
It's like saying that, all response times being equal, it is a fast application server.
What does that really tell you?
Python is also not all that slow, CPython is slow-ish but it's has effectively no optimizations at all. PyPy which has decent optimizations is actually fairly fast and getting faster with every release.
That being said, "right" is very different from pg's "powerful", especially as he says:
> Lisp is so great not because of some magic quality visible only to devotees, but because it is simply the most powerful language available.
In my opinion (and this person's[1]), Lisp is "powerful" in that it lets you express anything quickly and easily in a way that makes sense to you.
That doesn't mean it makes sense to anyone, nor does that mean it's compatible with anything else.
And where he says:
> What's less often understood is that there is a more general principle here: that if you have a choice of several languages, it is, all other things being equal, a mistake to program in anything but the most powerful one.
I feel he is sorely mistaken. The best programming language for the job is the one that most easily and quickly fits the requirements, given your experience level and the intrinsic attributes of the languages in question. That may be less true when you're working on a personal project rather than one for work... but from an engineering standpoint, development speed is very important.
PHP may be near-universally reviled as a language, but I won't contest the fact that it runs anywhere, quickly. If you needed to quickly put up a blog at least some time ago, the answer was almost always "use Wordpress and PHP".
Python is slow. Painfully slow. Slow as in "it stores the entire AST at runtime and lets you modify it while the program runs" slow. But it's also great at numerical computation with numpy. And it's enormously flexible (which is what makes it slow). And it's clean. It writes really well. For a quick mathematical processing program or website whose code needs to be legible, it's great.
This goes on for just about any language. There are reasons to, and not to, use a programming language. It's not just "power" that's important, it's a number of factors. I hate many languages with a passion, but I'll agree that those languages still have their uses. So I disagree with pg here.
1: http://www.lambdassociates.org/blog/bipolar.htm