Not by themselves. Spectral sensitivity curves let you compare how each sensor responds at particular wavelengths and how much the RGB channels overlap, so they can hint at color-reproduction potential. But they do not tell you overall color accuracy on their own.

Final color depends on the whole imaging chain: lens transmission, filters, sensor color mask/sensitivity, analog-to-digital conversion, in-camera color processing, and any later editing or output. Even if you ignore some of those steps, different cameras can still process the same sensor data differently.

So from the curves alone, the safe conclusion is limited: you can compare relative response at the same wavelength and see differences in channel separation/overlap. You cannot reliably say one camera is "more accurate" in color reproduction just from those plots.

To judge real color accuracy, you need a proper controlled test and evaluation workflow, ideally using calibrated output/display conditions.