Yev here -> while the numbers aren't entirely correct in our case the equation is correct. We do account for engineering time and opportunity cost of other projects that aren't getting worked on (though we tend to stack projects so while one team works on, billing, another will work on UI, when billing team is done they'll move on to the next project, etc..). One of the side-benefits for us in this case was building out the Extensions program meant we had to build a more robust crediting system, which is code we might be able to re-use later!