The focus of capital theory is once more on the macroeconomic aggregrate production function, almost seventy years after Joan Robinson launched her attack on this concept. The fundamental problem of capital theory is: Can techniques be so ordered and chosen according to their capital intensity that full employment results because of flexible real wages? I here return to the relatively new proposition that such an ordering is not only not possible, because it will occasionally be upset by reverse capital deepening. There is also the defficulty that substitution may simply not take place because only one, the optimal technique, will dominate over a wide range of change of the distributional variables. The conditions, under which such limitations to substitution hold, are not yet fully known, but we here offer a theorem that ensures such dominance under reasonable conditions.