> Actually, it is not meaningless, since I am trying to show the limits of the observation that in a balanced system small changes can have big effects.

You may be trying, but you're not succeeding. A single counter-example can never show a limit to anything.

Try applying your own example to the body heat of a person. For a person with a body temperature of 36 degrees celsius sitting in a 20 degree celsius room, you have 293 kelvins natural temperature, with an additional 16 kelvins "man-made" temperature. An additional increase of 2 kelvins will make you sick, and 5 kelvins will kill you.