Why do the estimated births/deaths per second counters have so much flicker? Surely you don't actually believe that the expected number of births/deaths per second fluctuates at 1dp precision multiple times per second?
Ok, so you added high-frequency random noise to the estimated averages to make it feel more realistic. To me, this makes it feel less realistic.
Anyway, don't mean to gripe, this is a cool project!
I think the more interesting fluctuations are those which change on an hourly ("checking the clock", spikes when people hear a chime or get notifications on the hour), daily ("eating lunch" spikes when UTC+8 hits lunchtime in eastern China and craters when it's noon in the middle of the Pacific), or other periodic basis.