I am basing my response to this upon the reading of two books, Black Swan by Nassim Taleb, and Stumbling Upon Happiness by Daniel Gilbert. I believe the combined knowledge of these books explains why trends are hard to predict.
Firstly, the future is tough to predict, its just too complex. There are varying degrees of prediction, and the more and more specific one wants to get, the more exponential the possibility of being wrong. It is easy to agree with my prediction that the sun will rise tomorrow [there is the remote possibility that even that would be wrong]. Me saying that the sun will rise tomorrow at 6:47am, with a mean temperature of 76 degrees a cross wind of 4 mph and a butterfly passing my window with blue wings, is a tougher bet.
Secondly, we don't know what makes us happy. When we think about the future and what will make us happy, we tend to localize into small portions of that future. Having that big house with the picket fence and well manicured lawn is what makes us happy right? What about the water bill, mowing the lawn, painting the fence, and the mortgage? There are so many other factors around the future that we filter out when making envisioning the future.