Think of it this way: infix syntax is just another form of DWIM. The computer is trying to act like it thinks the way you do. However, it doesn't, and because it's pretending to, bugs happen. Lisp doesn't pretend, and this helps to avoid bugs.
Infix is such a small percentage of what I do, it hardly matters. And where it's extremely critical, I have to share with non-programmer humans so consistency there helps.
>I have to share with non-programmer humans
Non-programmers read your code?
Why would it be harder to learn yet easier to read? If anything it's syntax is easier to learn (because there is so little syntax) but harder to read (also because their is so little syntax) and harder to understand (because the syntax doesn't reveal as much semantics).
There is a simple elegance to it but I also think it pushes a lot of complexity into other places. My opinion is that the trade off in where the complexity is doesn't benefit most programmers/projects -- Lisp users would obviously disagree.