Yes, evidence is very important. However, it is important to START with the evidence and allow it to lead you where it will. If you START with a pre-conceived notion (such as "explosives were used in the WTC") then the evidence could lead you astray.
Here's a hypothetical example: Supposing I believe that a giant purple salamander is going out every morning and eating all the newspapers off the neighbor's driveways. I want to test this hypothesis by going out at 5:30 AM, right after they are delivered, and confirming that the newspapers are in fact lying on the driveways. An hour later, I go check them again. If my hypothesis is true, then the newspapers will be gone. Thus, the purple salamander hypothesis is proven.
See the problem here? I am so focused on my pre-conceived belief about what is happening, I fail to consider other, more mundane explanations.
That is why falsifiability is so important. If the purple salamander hypothesis were wrong, how would I know? Certainly not by watching the newspapers and finding that, in fact, the neighbors themselves are picking them up and taking them into their houses. I could simply say, "They beat the purple salamander to it", or "the purple salamander was sick today".