In theory you can numerically determine the intermediate value that gives the largest or smallest bound by using numerical evaluations of functions and possibly their derivatives and solving a combination of optimization problems based on numerical function and derivative values. However finding the intermediate point that gives a bound with a formula will involve solving an optimization problem that involves combining a polynomial and a possibly very complex e.g. transcendental function, and in general those kinds of optimization problems cannot be solved analytically with a closed form formula.