Your main advantage will be to take the requirement and break it down into its lowest parts and then hand off those parts to an outsourced developer and if he/she does not deliver exactly to your spec, then end the relationship. Period.

There is a certain amount of luck to it, but just as the author suggests, I recommend first having the developer create a very simple application with requirements that you define and estimate how long it would have taken you (probably no more than a 5 hour task) and if he/she can deliver, then you've got yourself a winner and you can focus on the business and not javascript errors.

There is a fine line between clarifying and micromanaging though.

Initially I start by posting the job, then filtering in stages before asking them (and paying them) to complete a small 1 to 2 hour coding project. The coding exercise can be done many different ways but the purpose is to evaluate problem solving, code quality, knowledge of libraries, code structure, etc. It's amazing how much insight you can gain with something like that.

More importantly, having them work on something real helps me further evaluate communication, responsiveness, and other attributes I look for in a good dev.

The rewards of finding the right developer far outweight the upfront cost of looking for one in my opinion, but it depends how long you are planning to be in business for.