Market.
Viaweb and YC were each aimed at the overlooked low end of a market. Small businesses may not be sexy, but they had a pressing need for a web presence, and the can't write it themselves.. College students and young professionals may not be experienced, but they have a real need for money and advice, and they (usually) can't fund it themselves.
Arc, however, is explicitly aimed at the high end. It's meant to be a "LFSP". And smart people can write Arc themselves (and have ;-)). They don't need a new programming language; if they wanted one, they could invent it themselves.
Hence the contemptuous reaction, and why I think Arc is in more trouble than either Viaweb or YC. It's not really a problem when your peers are contemptuous, because that just means they're less likely to compete with you. It's a big problem when your users are contemptuous, because that means they don't need your project.
If I were in charge of Arc marketing, I'd position it as the "PHP for the ones that PHP forgot". PHP initially got its start as a way to throw a quick & dirty webapp prototype up on the web, and then incrementally refine it. But around version 5, it started getting complex and adding all these features from Java. That's left a vacuum at the bottom of the web language market. Rails and Django tried to fill it, but the average Django user is quite a bit more sophisticated than the average PHP user circa 1999. And Arc's design is already pretty well-suited to throwing a design up on the screen quickly.
Right now, the majority of Arc users seem to be disenchanted Lispers. Disenchanted Lispers tend to be smart people, and they also tend to be interested in language design. So instead of doing things with the language, they do things to the language. On arclanguage.com, I've seen one person write an app with the language, and several dozen people write enhancements of the language itself, either in the form of macros or as hacks to the interpreter itself.