There is scope for considerable refinement in the choice of the second factor p(life|HZ). [...] In our simplified treatment, this constant value has no effect o

There is scope for considerable refinement in the choice of the second factor p(life|HZ). [...] In our simplified treatment, this constant value has no effect on dP(t)/dt since its contribution is also cancelled by the normalization factor N. Though this paper does provide some needed formalism around the Drake Equation, I don't understand how its results can be meaningful if it "cancels out" one of the largest sources of uncertainty.