Since there are examples of locations having a more accelerated rise over time in tooth decay when water fluoridation is discontinued [1], I think one can extract from that, that possibly the explanation is that enough water is being drank/used in ways where it contacts teeth, that a difference is clearly seen in the dental statistics of a population monitored over time when fluoridation changes are made.

[1] Article/study discussing Calgary vs Edmonton in Canada, linked elsewhere in the comments here that has this snippet:

What’s happened since 2012? A study found that within a couple of years of ending fluoridation, decay in children’s baby teeth had risen by 65 per cent compared to 2005, while the increase was only 14 per cent in Edmonton, which continues to fluoridate its water.

--

I won't argue the fact that there may indeed be problems with ingestion, and additionally that contact to the teeth is the only way for the fluoridated water to help, but I'm not convinced that existing water fluoridation isn't having a positive effect on tooth decay. Whether it's worth other potential health trade-offs is certainly a valid concern.