The ability to now make measurements of Be and B as well as put constraints on lisix abundances in metal-poor stars has led to a detailed reexamination of Big Bang Nucleosynthesis in the $Agroughly6$ regime. The nuclear reaction network has been significantly expanded with many new rates added. It is demonstrated that although a number of $A>7$ reaction rates are poorly determined, even with extreme values chosen, the standard homogeneous model is unable to produce significant yields (Be/H and B/H $<10^{-17}$ when $Ale7$ abundances fit) above $A=7$ and the liseven/lisix ratio always exceeds 500. We also preliminarily explore inhomogeneous models, such as those inspired by a first order quark-hadron phase transition, where regions with high neutron/proton ratios can allow some leakage up to $A>7$. However models that fit the $Ale7$ abundances still seem to have difficulty in obtaining significant $A>7$ yields.