1/ They exist so US gov't agencies can get products they need developed, especially when the risk involved in R&D means they judged not worthy of private investment.
2/ First round SBIR has somewhere <20% acceptance rate, so you'll need to plan to make multiple applications over multiple periods. The effort isn't less than the VC route.
3/ Develop a relationship with the sponsor (person who put out the solicitation). This is similar to networking with VCs and prospective customers.
Working with the US Government is a very different set of problems than VC money.
Actually, performing work under contract with any customer is very different than taking VC money. With VCs, it's understood that it's a risky investment and the company that's being funded can fail before it delivers a product. With a contract, you have specific deliverables (sometimes at a fixed cost), and the other side can sue you to recover their money if you don't deliver. (Under certain circumstances, your creditors can even go after your personal assets.[1])
Also, in an SBIR, the government agency gives you funding to research and/or develop a specific thing that they need, so you can't arbitrarily "pivot" to an unrelated thing if the first thing isn't going well.
[1] https://en.wikipedia.org/wiki/Piercing_the_corporate_veil#Un...