Part of the article's suggestion is to better differentiate monitoring points, so as to divide farm-related hits from urban hits.
That said, in cities that have combined storm and sewer system, the latter is a lot more difficult.
But in general, it is a uniquely politically-palatable continuous monitoring solution.
The public doesn't want to think about what happens underneath a toilet or drain, so no one cares if someone is sampling and testing it.
SF has a combined sewer system, but it has zero farms. One question is whether it was more likely to be a signal from wild birds (into the runoff system) than humans (into the sewer.)
There's very little rain in the summer months, the average for July rounds off to 0.0in. I don't remember if there was any significant rain in May or June, but I'd be surprised if it was more than 1-5% of the total. I guess there's lawn irrigation, though.
SF has pretty limited lawn irrigation as well. Depending on how the sewerage system is divided, some parts of the city have effectively none.
Also, it's not so finely-grained that it seems like a privacy problem.