In previous work, we showed that there is indeed a fundamental limit to the off-resonant nonlinear susceptibility in the dipole approximation. This limit is calculated using sum rules. We stress that there are no approximations used in deriving the sum rules and they are equivalent to the Schroedinger equation. As such, the sum rules must be obeyed if quantum mechanics is correct. Over the years, complex quantum expressions have been simplified with two and three state models that consider the competition between excited states, symmetry, and bond length alternation. We can show that while such models are good approximations to the measured susceptibilities, since the matrix elements and energy levels may not be independently adjusted, they serve limited usefulness in understanding when the susceptibility is maximum. A large set of measurements form the literature shows that all molecules fall below the susceptibility horizon calculated buy many are close. Further improvements in susceptibilities can be made with the sum rules as a guide. Significant additional improvements will require radically more creative approaches than are presently used.