National enhanced rock weathering (ERW) maps can overstate decision confidence when environmental heterogeneity, uncertain geological evidence, and modeling assumptions are compressed into a single suitability score. This study developed the Ghana Rock Enhanced Weathering Opportunity Index (GREW-OI), a 2.5-arc-min reconnaissance framework that separates hydroclimatic opportunity, acid-soil co-benefit, and mapped candidate-lithology proximity and integrates them using a weighted geometric mean. Its methodological advance is to evaluate parameter sensitivity, geological structural sensitivity, soil prediction support, and district reporting support separately, rather than combining them into a single uncertainty estimate. A total of 18 modeling assumptions were explored using a 3,200-member Latin-hypercube ensemble, with geological classification and soil prediction support assessed independently through deterministic sensitivity brackets. Across 10,680 complete cells, 4.76% exceeded the illustrative score threshold of 70. Parameter variation produced a median cell-level P90–P10 width of 10.30 points, compared with median score ranges of 12.06–18.88 points across soil-support brackets and 19.83 points under geological half-class perturbations. Geological classification produced the largest model response and materially altered the mapped decision space: the fraction of cells exceeding 70 ranged from 0.54 to 15.28%, while the number of districts reporting at least one observed exceedance ranged from 6 to 85. The GREW-OI reframes national ERW mapping as an evidence-prioritization problem, identifying where improved geological, feedstock, soil, land, logistics, and field evidence would most effectively strengthen carbon dioxide removal decisions without equating mapped opportunity with deployable areas or net CO2 removal.