Low test weight does not cause lower yield.  Low test weight would/might be an indicator that your yield will be lower, but it does not cause yields to be lower. Look at it this way..... the article made the point that we really deal in "56-lb chunks" for corn (and "60-lb chunks" for soybeans) which we call bushels.  In actuality, you are hauling weight - lbs & tons - to the mill, not volume - bushels.  The only reason bushels matter is that we - the USA and the English system - price grains by the bushel.  If you lived in Canada or Australia, where their grain is priced by the metric tonne, we would not be having this conversation.

In reality, the weather conditions (and other factors such as genetics/ hybrid chosen) will affect the lbs harvested, and those lbs harvested are your measure of yield. 

Question that crossed my mind as I read the article that you linked: it says that breeders like to measure test weight because it is easy to do.  So, my question is:  if weather impacts test weight, how accurate is that measurement?  Wouldn't it be conceivable that hybrid A tested 58 lb/bu in 2008, and 52 lb/bu in 2009, and that is simply a reflection of weather conditions?  Is it possible that hybrid A will test 58 lb/bu on Farmer Jones farm, where he does well in managing that hybrid, and only 56 lb/bu on Farmer Smith's farm, who does not do as well in managing hybrids?  I would think these statements are true.  So, plant breeders might get an general idea of how a hybrid will tend to test, but they can't accurately predict it, because they can't accurately predict an individual farmers management, or the weather.   Am I correct, or am I all wet?