Hello Conrad, this is as well a preference :)
I personally believe weighting can give very useful results. When you take only normalized results, you are also applying a default weighting, that dictates that each impact category is equally important. I do not agree with this "equal weighting" apporach to normalized scores, personally.
There are many arguments for and against each weighting method, but i believe weighting certainly has its uses, and should not be discounted as a tool.
But also yes, the PEF weighting has a lot going against it. I hope they will adress some of the more serious concerns in a new iteration of weighting factors.