The very famous meta studies with all the negative correlations get all the bad associations with flouride from regions where water naturally has extremely high (relative to most other parts of the world) levels of fluoride in addition to high levels of many other uncommon concentrations.
Some of these regions also have additional problems with industry waste.
Put simply, negative correlations about unattended children in swimming pools cannot be extrapolated to infer negative correlations about young children and sippy cups of water.