A possible loophole is that in the nuclear setting, the total amount of negative charge is the same when considering the eletric monopole shift, while in this toy model, the charge in both situations is different, being -2e in toy model 0 and -(2e+epsilon) in toy model A.