This post relates directly to the previous post. I am not going to tell you why; I'd build you a rifle but not shoot for you, I will build you a fire but I will not cook for you; I will present you this analogy but not explain it to you.
So, building on the last post and also going directly from the Meta-Chanics post, I'm going to start this 1-of-2 post, about testing pure systems.
So, when you're playtesting you need to know what it is you're testing. If you're saying "System" then more likely than not you are deluding yourself, for you are testing the Game as a whole. I believe much can be learnt from testing the system on its own.
So how does one test the system on its own? First, one cuts off all mentions of colour and background. The three Aspects become Three Aspects that cover A, B and C, or even just three Aspects. This is easy, and allows you to see what directions it could go on once the colour is peeled off, which happens with most games after a while anyway.
Then, you take away all mentions of how the game should be played, the "Meta-Chanics", you don't tell players if they should go cooperatively or competitively, you let their own nature take them where it will.
This is the way to see what the System encourages, rather than what you encourage. If you want your design goals and what the system actually encourages, rather than what it encourages once you point it in the "right" direction, then you have to see what the System does.
The System and not the Game.
And you could see it as a way of me agreeing with Troy, Setting does matter. Though I look at it from another angle.