Original equations contain four parameters $b,d,s,e$ so that in order to discuss the behavior of dynamics, you have to fix three parameters and investigate the case varying the rest one. That's so complicated of course. After introducing $h,p,\tau,\rho$, equations contain only one parameter $\rho$ so that in $(h,p;\tau)$ space, no matter how $b,d,s,e$ change, which means you don't have to fix anything, if we know the behavior of dynamics varying $\rho$, everything seems clear.
Let me give a more intuitive example. I want to investigate the shape of the region $x^2+y^2\leq1$. Instead of fix each $x$ and find $y$, we set $r^2=x^2+y^2$ and the region becomes $r^2\leq1$, in which case we don't need to consider both $x$ and $y$ and instead focus on $r$. When solving equations, variables substitution usually gives simpler view.