Your result is including all infractions of any type across the board and indicates that poor water safety is rare and isolated.
To get the total sum, it's compiled over 34 years, and says: "relatively few community water systems (3–10%) incur health-based violations in a given year."
It also states that in the most dense parts of the country, urban regions, there are rarely problems.
As well as this conclusion: "Generally, water systems in the United States provide reliable and high-quality drinking water. Violations tend to be infrequent."
I already covered that quoted segment. The 21 million figure includes all infractions of any type, regardless of severity.
The primary problem in the US is aging infrastructure in sparsely populated rural areas, as stated in the study the USA Today article quotes from. Nearly the entire US gets a good violations score in terms of low instance counts, except for Oklahoma (they show an extreme number of violations), rural West Texas (also intense), and a few isolated rural pockets in eg California and Idaho (~1/10th or fewer the violation counts of the OK & W.TX examples). The OK and West TX examples are no doubt caused by the polluting energy industries there.
You also have to look at the standards themselves. If a US water supplier violates a standard about some chemical not having more than 2 parts per billion because it had 2.5, and in some other country there's no violation because the standard is 15 parts per billion and they had 12, that's still not an advantage to the other country.
Last time I looked at this, yes, the US has very high standards and tends to meet them. Western Europe does as well, and also tends to meet them. ISTR there wasn't a systematic difference between the two, either; this chemical's limit might be slightly high here, but some other chemical's limit would be slightly lower. It looked more like what you'd expect from two similarly-concerned regulators that simply have no reason to coordinate than either side having higher standards.
You can't just look at violations. That procedure always makes things look bad because even just one violation is a violation, even just one error is an error, oh no! You have to look at the system holistically, and, yes, US water is pretty darned good when you look at it that way. You can't just look at problems but pretend to yourself or others that you're looking holistically at the situation.
(The Internet trains people to go "Oh, so we should just stop worrying about it then?" when someone says it's going pretty well, but what I'm actually saying is that our current level of concern is pretty much correct. It shouldn't be lowered. But there isn't much compelling reason to raise it globally, either. Only locally in certain places in response to local issues.)
To get the total sum, it's compiled over 34 years, and says: "relatively few community water systems (3–10%) incur health-based violations in a given year."
It also states that in the most dense parts of the country, urban regions, there are rarely problems.
As well as this conclusion: "Generally, water systems in the United States provide reliable and high-quality drinking water. Violations tend to be infrequent."