I don't follow your argument though, we have a continuous knapsack problem.  We want to optimise total cost (fossil fuel usage) by investing in X or Y.  X has a lower cost than Y, so we should invest in Y. (really we need to know the dX/dcost vs dY/dcost and pick the greater)

Please note, it was someone else commenting on on his blog:
https://www.blogger.com/comment.g?blogID=27481991&postID=2232157439199671445