The CoffeeScript syntax is nicely readable, but IMO the C-like syntax is far more easily 'scannable'.
Quickly scanning over the C-like syntax I can more easily notice the ! and the && than I can make out the 'not' and the 'and' while scanning the more natural-language-like CoffeeScript because of the clear separation between ! and && and the alphanumerics commonly used in variable names. IMO the 'and' and 'not' keywords kind of get lost in the noise of all the other characters.
YMMV here. I've been programming in C-like languages for quite a long time so maybe that's rewired my brain in a way that doesn't hold for people who don't program in C-like languages.
