If we look at the "startup industry" as an industry, we should ask ourselves some tough questions like... What other industry in the world has an 80% to 90% failure rate? And why is such a high failure rate acceptable?
I don't fault Startup Chile for their program and think we should all be working to push these startup conversations forward. To improve the odds of a positive outcome, here are the questions I would ask if I ran such a program. "How much more effective would these programs be if the focus was on agreed to customer-centric value creation milestones and then capital was deployed based on achieving these milestones? What if the entrepreneur or startups were measured on building products that 1) nail the customer's pain, 2) nail the minimum feature set to satisfy the customer demand, 3) nail the business model or go-to-market strategy? etc...
In addition to reimbursing for expenses, an effective startup program could tranche additional seed capital based on how the startup achieved early success and nailed early value-creation milestones. Give high-potential entrepreneurs a little bit of guaranteed runway is great. In addition, provide them follow-on financing in the form of tranched performance payments based on successful execution.
Doing this would require an agreed to framework. Here's a link to a recent conversation thread on one possible way to achieve this. http://news.ycombinator.com/item?id=3406330