Sandia Labs did a second 1/4 scale test with heated Jet A, the full report can be read here, and stated that:
behavior. This study was not sufficient to establish a fundamental understanding of the mechanism of quench behavior. In light of the small number of experiments, modeling of the full scale has to be regarded with a great deal of uncertainty. At its current state, modeling can not provide a definitive tool in determining the location of the ignition in the CWT. The combustion behavior of actual jet fuel mixtures is far more complex than originally envisioned. The lack of definitive initial conditions and the weak basis of the current understanding of combustion behavior in multicomponent fuel-air mixtures lead to much uncertainty in the modeling.
In other words, they couldn't replicate the CWT explosion of TWA 800.
As I recall the “Mythbusters” test, they used not only the conditions there, but LESSER conditions (pressure/simulated altitutde) and still got a massive explosion).