Well at this point we're getting away from my area of knowledge (the neurophysiological causes of depressive illness) and into clinical psychology and diagnosis. Therefore I can only make an educated speculation but I'll give it a go.

I believe part of the problem lies in thinking of depression as something you either have or haven't got. I believe it is more of a spectrum disorder like Autism. Let me explain.

Broadly speaking, a depressed state is caused by a decrease in the action of serotonin in the central nervous system for whatever reason. In other words, just feeling a little bit sad can be attributed to this - in other words if you're feeling a bit down in the dumps one day, you can be described as being depressed. Your mood and mental functions are depressed due to (among other things) a decrease in the action of serotonin.

Now imagine the more extreme version. You're really really sad, struggling to think of anything positive etc. I believe that this can be put on the same scale as the "feeling a bit sad" state - this is sadness turned up to 11, and is caused by a major lack of serotonin and/or its actions. I mentioned MDMA before, and it can serve as an example here. During a comedown, when central serotonin is depleted, you feel like shit. Not quite as bad as during major depression, but along similar lines. On the other hand, during the come-up, when you're full of it, you feel amazing. Manic.

What I'm trying to say (in an admittedly roundabout way) is that I think of mood as a scale from, say, -10 to 10. 10 being bouncing off the walls manic and -10 being suicidal. Everybody has the capability to feel 10 or -10 and everything in-between (because in theory everyone can have loads of serotonin or none), but some are pre-disposed to it, and some are more at risk of it from risk factors like stress, but most people range from, say, -5 to 5 during their life. People who "suffer from depression" in a clinical sense just feel abnormally depressed more of the time and more easily.

Anywho, on to your question which was how to test effectively for depression (if we take depression to mean feeling sad an abnormally large amount). Firstly, the benefit of such a test would have to be to recognise those feeling sad enough for it to impact significantly on their lives. Thus, I believe the Beck Inventory is sufficient - anyone who is depressed enough to have it actually affect their life will score highly, anyone who's a bit sad will score low.

Another thing to do (although it would be impractical) is to do a spinal tap and measure the amount of serotonin in the Cerebro Spinal Fluid (the fluid that bathes the brain and spinal cord). This would give a good indication of whether a person's serotonin levels were too low compared to a statistically determined mean. However, limiting serotonin's actions (and thus causing depression) can be done in other ways that this investigation would not allow for. For instance, overactive reuptake channels that take serotonin back into nerve cells, overactive action from enzymes that break serotonin down, lack of serotonin receptors on the synapse.

My point is, you can measure serotonin levels and enzyme activity until you're blue in the face, but the best way to tell if someone has a problem with depression is to do a behavioural test because depression, if it is bad enough, is a behavioural illness.