Recent advances in 3D model atmospheres have led to a significant downward revision in photospheric heavy element abundance estimates for the Sun, especially for the abundant CNO species. However, adopting such low abundances leads to serious conflicts with helioseismic data. I argue that there are significant systematics in surface abundance measurements, and advocate a interiors-based approach similar to that used in the solar neutrino problem. In this method, meteoritic and helioseismic data can be combined with interiors theory to predict absolute solar abundances which can be used as a reference for atmospheres models. I discuss three distinct seismic methods and solar neutrino fluxes which agree on a high solar metal abundance, and constraints on neon from the sound speed profile. I will also present evidence in the sound speed profile for mild internal mixing, consistent with the solar lithium depletion, and discuss experiments in progress to measure opacities under stellar interiors conditions.