Why not CL?
What made me switch was that Lisp machines (both Symbolics and LMI) were so gratuitously, baroquely complex. The manuals filled a whole shelf. Each component of the software was written as if it had to have every possible feature. The hackers who wrote it were the smartest and most energetic around. But there was no Steve Jobs to tell them "No, this is too complex." So the guy in charge of writing the pretty-printer, for example, would decide. "This is going to be the most powerful pretty-printer ever written. It's going to be able to do everything!"
Unfortunately this complexity persists in Common Lisp, which was pretty much copied directly from ZetaLisp. In fact, both of the worst flaws in CL are due to its origins on Lisp machines: both its complexity and the way it's cut off from the OS.
Indeed, this is one of the reasons why Common Lisp is much less appealing to me than Scheme. I mean, I love the idea of a Lisp machine. I loved reading the old Zetalisp manuals (long after they were relevant -- I wasn't around back then). But I don't love Common Lisp, for several reasons, and one of them is that has accumulated too many features. (There are other reasons, but that's something for a separate post. So are my reasons for choosing Scheme, and choosing Chicken specifically.)
Having many features doesn't make something a bad language, but it's the "everything and the kitchen sink" mentality that turns me off. It reminds me of languages like Perl, or Ada. :-/ It's just not something I am looking for right now.
(It's an odd comment coming from PG though, considering he has written books about Common Lisp... but maybe this is one of the reasons he decided to start writing Arc.)