You have to look at the figures in more detail before making such assertions, and take into account that the resistance of the joint is initially very small compared to that of the load. If you draw a graph of power dissipation in one of 2 series resistances (joint plus load) at constant supply voltage, you will see that it peaks at the point where the two resistors are equal, so until the joint resistance equals that of the load (by which time you are in serious trouble), power dissipation, and therefore heat, will increase with increasing joint resistance.