For ethanol precipitations, as mentioned in https://en.wikipedia.org/wiki/Ethanol_precipitation, given that the attraction of water molecules to charged ions like phosphates is based on Coulombic attraction, shouldn't the water-phosphate attraction be increased in magnitude by the same ratio as that by which the phosphate-cation in increased in magnitude? I.e., should there not be a lack of increased "preference" for phosphate-salt cation ionic bonds (and thus precipitation) over water-phosphate hydrogen bonds (and thus polynucleic acid solvation) as a result of the increased permittivity to charge-charge interactions that ethanol brings over water? An argument based on the relative higher amounts of cations and ethanol, and lower in water, would make sense in increasing the likelihood of a cation-phosphate bond over a water-phosphate hydrogen bond, but then any mention of the dielectric constant in explaining ethanol precipitation seems entirely superfluous.
The interaction between negatively charged nucleic acid and positively charged counter ions is an ion:ion interaction. The interaction between ions and water is an ion:dipole interaction.
The presence of water between ions of opposite charge diminishes their long-range attraction (solvated ions are "shielding" by the oriented dipoles of the solvent). Ion:dipole interactions are short range and rely on the dipole to be properly oriented. Adding ethanol to the aqueous DNA solution does not make the ion:dipole interactions of water with ions stronger. In contrast, it does make the ionic interactions stronger because ethanol is less effective in shielding charges than water.
It is important to compare dielectric constants of water and ethanol because both molecules are uncharged and polar (and capable of ion:dipole interactions), but water is the better solvent for ionic compounds for the reasons discussed above.