My two cents worth... I think these estimates are not conservative enough, I think it is better to approach this "semi"-modeling exercise (pun intended) conservatively rather than tending towards overestimating - if the GF3 happens sooner, if the CT is pumped out quicker on the line, if less...