If you're an engineer, it's a bad idea to join a start-up if their focus doesn't align with your interests. Even if there is significant equity and a great salary (it's not uncommon for funded start-ups to pay more than big companies), it's terrible to feel underused, unfulfilled and behind the technology curve in the topics that interest you. If you're the _only_ person knowledgeable about a certain topic in a start-up, there's nothing you can learn about it from coworkers. If one is uncertain of their focus, there's also no ability to switch groups and work on something else.
Your start-up (OpenDNS) is fairly unique in San Francisco, if not Bay Area as a whole, in being focused on network infrastructure (and I applaud you for it, it's a great service!)
Unfortunately, it (+ several other start-ups e.g., rethinkdb) is an exception, not the rule. The focus of most other early stage start-ups is web development. Engineers interested in solving other kinds of problems (e.g., algorithms and scalability challenges) but who like the flexibility of working for an Internet (vs. shrink-wrap product) company typically consider the medium (Facebook, LinkedIn) and big players (Google, Yahoo, Amazon).
Some people do best at the 1-5 person company. Some people do best at the 5-25 person company. Some people do best at the 25-100 person company. And some people do best at 100-250, and 250+.
The best people are the ones that grow with you as the company grows and take on bigger leadership roles as needed, constantly working for what's best for the company and it's people.
The people Google is hiring at 20,000 are generally not the people I want as we go from 25 to 100. That's not universally true, but when we lose a candidate (which is rare) it's because they go to another start-up and they had to make a real tough choice, often resorting to commute and other lesser factors to decide. It's not because they went to Google or LinkedIn.