Because nothing will be in that water that wasn't already in there from the treatment center, except maybe some extra copper if they are using copper heatpipes or something. There's nothing additional to treat, and nothing to be any more harmful on reintroduction than it was already. If it's something they add, it's something they monitor. The only things that would accumulate from halving and reintroducing the water into the system is stuff they would add, and thus monitor.
Essentially, in the scenario listed by the post I was replying to:
Water with chemicals x y and z is provided by the water center.
The water evaporates, leaving concentrated doses of say, x and y, but z evaporates.
Water with high doses of x and y are undrinkable.
I say, it doesn't matter. Since x, y, and z are monitored.
Highly concentrated water will be introduced into the water, and will be tested for. The dosage will thus be adjusted accordingly. Those random chemicals you're referring to are irrelevant because they aren't arbitrarily added by the water being processed at a waste water treatment center. Therefore, they aren't going to be any more concentrated than they would have been before. If anything, they are added by us pissing whatever drugs we've taken into our toilets. (which water from a data center is unlikely to have any more of than it did already.)
You can't even really complain about the frequency of testing because the load of the data center is going to be pretty constant. Consistent levels of chloramine, in my example, will be returned to the treatment center. (Btw, chloramine is Chlorine and Ammonia, and is used in place of chlorine to prevent bad stuff from growing in the water.) This stuff is used because it is more stable than chlorine and will /not/ evaporate. It will be stuff like this that survives the return trip into the water system at concentrated levels.
Maybe I am missing something about waste water treatment, and what chemicals are involved, but meh.