Sure you can throw together another Snapchat in a weekend. But very soon - if you have real significant user base - you will need to build and maintain apps on all platforms. You will need someone to fix bugs, add new features and scale things better everywhere. You will need people who can create and conduct marketing campaigns (FYI - Techcrunch isn't writing about for you without getting paid and Snapchat didn't become popular without playing tons of marketing tricks). You will need guys who can do business development, accounting & taxes, create partner relationships and so on. You will also need customer support, social media management, metrics measurement, all kind of dashboards and so on. Before you know, you will end up with significant staff, needing office space, doing moral events and likes.

Per head aggregate cost of an engineer these days is upwards of $230K when you add up all health insurance, 401K, signing bonuses, office space and so on. Even if you hire just 50 employees, you can easily expect $1 million burn rate per month. For things like Snapchat, you can't anticipate much income during initial growth period. So this means you need VC for getting ~$10-20M of funding per year until you get sold, IPOed or have ad based income flowing in.

In support of your thesis, there was actually one guy who did plantyoffish.com and managed it through out everything all by himself for many years. But finally things have caught up to him. Now he has more than dozen employees. This is of course outlier case and not a typical scenario where competition will eat you alive if you don't defend your castle aggressively every day by one-uping them. And for that you need employees and bandwidth to work on lot of things at the same time, many of which likely won't work out.