But your conclusion is disturbing:
Probably the poisons that would set off alarm bells most are rather hard to get.
"Hard to get" in the sense of being expensive? Well, somewhere, someone with a lot of _____ money would probably be willing to fund a "research" project. A well-coordinated, large scale effort could prove to be a cost-effective way to eliminate one's enemies.
Secondly, if the contaminates are introduced at the user side of the filtration, how much damage will occur before any such alarms are activated? (Rhetorical)
Hope our people are all over this.
When I say hard to get, I mean it. As in there is no retail of such substances, and big industrial plants have to pay through the nose for it or make it themselves. It is the furthest thing from “cost effective terrorism”. More trouble than it is worth.
Probably the worst natural poison for this sort of thing would be arsenic, and to say that the EPA and the USGS are obsessive about it is an understatement. Since 2001 the most they will tolerate in drinking water is 10 micrograms per liter. 10 millionths of a gram per liter.
http://water.usgs.gov/nawqa/trace/arsenic/
And yes, every water treatment plant in the US constantly tests for it. If they get so much as 11 micrograms, they dump drums of iron oxide into the water, which binds with the arsenic and precipitates it out.