What I meant was, Lisp programmers usually learn to ignore the parens. When I look at a Lisp code, I just see indentation and code. Your eyes train themselves to do that. Also I have never found excessive parens to be an issue at all. But like I said, maybe thats just me.
Clojure was alien. When extrapolating what it might be like to work with Clojure, prefix notation (+ 1 2) was an obvious downgrade to infix notation 1 + 2 in a REPL. Arrow-keying your way around s-expressions on http://tryclj.com/ just seemed to confirm it.
But it's sort of like Vim: if you could see your Clojure workflow in a month, you'd be sold now. But since you can't, and since you don't know about Paredit, and since you don't realize that you probably will never even use a REPL since you'll be evaluating code in your actual source files, it's hard to be convinced and that's just how it is.
And there's still the risk of a month going by, just like in Vim, where you never even arrive at that killer could-be workflow because you never happened to stumble upon it and you don't even know it exists.
Most just look at the code from outside, frown at it and leave without ever knowing the immense power the language and its workflow gives you.