Understanding the detailed production and hadronization mechanisms for heavy quarkonia and their modification in a nuclear environment presents one of the major challenges in QCD. Calculations including nuclear-modified parton distribution functions (nPDFs) and fitting of break-up cross sections (sigma_breakup) as parameters have been successful at describing many features of J/psi modification in proton(deuteron)-nucleus collisions. In this paper, we extend these calculations to explore different geometric dependencies of the modification and confront them with new experimental results from the PHENIX experiment. We find that no combination of nPDFs and sigma_breakup, regardless of the nPDF parameter set and the assumed geometric dependence, can simultaneously describe the entire rapidity and centrality dependence of J/psi modifications in d+Au collisions at sqrt(s_NN) = 200 GeV. We also compare the data with coherence calculations and find them unable to describe the full rapidity and centrality dependence as well. We discuss how these calculations might be extended and further tested, in addition to discussing other physics mechanisms including initial-state parton energy loss.