I buy the argument that an auto-close horizon has value, but two months is far too short. More harm than good is done by aggressively closing topics, as it fragments knowledge and makes it hard to ask a question which is somewhat off the beaten path.
Consider this topic, Requesting Complete Example using build.mk. It was very helpful to be able to unite the question and the feedback. Otherwise, in order to document my findings I would have had to open a new topic, which only says "In reference to ..., here's something good to know". Obviously that's not good because on a long-enough horizon the search results become noisy to the point of meaninglessness.
Do you know who's in charge of setting the topic life? I'd like to pitch that person on dramatically extending it. If the goal is to keep 3-year-old convos from resurrecting, then maybe 1-2 years should be the horizon.