If I'm understanding you correctly, I dig it. Experimental sociology is a pain — if you control the environment by performing it in a lab, it's expensive and you probably introduce bias into the sample during recruitment. If you perform the experiment in the wild, it's cheaper, but you have even less control over what's going on.
Now, a whole town that unscrupulous social scientists could rent out for their own experiments (or just one particular group is doing it) by way of altering their environment or some people therein via anomaly? The utility of that is readily apparent, and if anomalies were real I could see it happening. That's pretty disturbing, and apart from that incident log you mentioned, you could fill the article out with descriptions of events that are based on actual experiments, or ones that people have wanted to perform. It'll take research, but it should be worth it.
it's worth noting that if you want to study how actual people act in a natural setting, you're not going to learn anything by altering everybody's behavior. Changing a whole town to be sadistic could be interesting for shits and giggles (if you were, you know, an awful person), but given that there aren't any real-life towns full of sadists to study, it would be kind of a pointless exercise scientifically. You get the idea though — the piece takes on a very different tone if the clock tower is just fucking with people to see what would happen vs. if it's setting up actually informative scenarios.
Another thing to think about is that something like this would be very useful to anyone interested in engineering a society from the ground up — the big problem with social planning of any sort is that you can't really test it out in a controlled environment first, and that's exactly what this clock tower could provide.