Yeah, you wouldn't divide the coefficients. You could calculate the value added by the next active user added to the network at the network's current sizes. So, perform both calculations for n active users and n + 1 active users and subtract. Then divide the results for both Steemit and Facebook. You would also need to make sure you are measuring against the same time span. The coefficient they used for Facebook was for monthly active users, while I just borrowed the daily active users number above.

But the issue here is that Steemit is so new that we don't know if we have a reliable trend against Metcalfe's Law like Bitcoin or Facebook currently have. So the biggest question would be finding a good way to derive that coefficient. The estimate I did as a lower bound might still be too optimistic given the small user base that we currently have. Metcalfe's Law might not kick in until we reach a certain point.