One thing that bothers me about TDD is that whenever anyone talks about it, its always such trivial examples as to be ridiculous for explaining how to do it. For example explaining TDD by using the example of addition. Seriously, testing adding two numbers. I personally think the reason for this is that testing is really difficult, maybe more difficult than the problem being tested. So the cognitive load is at least doubled, maybe more.
With this methodology you accept the results as "correct." We programmers do this anyway, we validate what we are working on in some fashion so we "know" what the correct results look like.