A recent attempt on my part to re-open an estimation question has failed, so I thought it might be worthwhile to continue the discussion on meta. The original question:
This was marked as off-topic since physicists should not care about the number of hairs on a human head. I will resist the urge to answer that physics = everything, since then we will start getting questions about the economy on here...but I think that estimation questions should be allowed for three main reasons:
1) Every single physics textbook contains them. If chapter 1 of an introductory physics textbook is "off-topic", then how can we determine that anything is on-topic?
2) Physicists have a long tradition of estimation. The phrase "Back-of-the-envelope" originated in the hard sciences, and any googling on the issue will immediately tell you these kinds of problems are referred to as "Fermi problems" because he was so good at them (http://en.wikipedia.org/wiki/Fermi_problem). He used them to solve physical problems - like estimating the size of nuclear explosions.
3) They are a key part of the working physicists toolkit. During seminars/conferences, questions of the form "it seems like that value is too large" are important discussion points, and originate with such estimation techniques. They allow us to study the mechanisms behind physical phenomena and processes which we may not be able to understand from first principles ( i.e. the Drake equation). If someone is running a machine (accelerator, telescope, SEM, SQUID...) they will be required to make on-the-fly decisions based on estimations of physical parameters of the system.
So can someone tell me why these questions are not allowed? To me they are really characteristic of problems in physics. I hate using quotes to prove arguments, but I'm going to do so in this case:
"Never make a calculation until you know the answer: make an estimate before every calculation, try a simple physical argument (symmetry! invariance! conservation!) before every derivation, guess the answer to every puzzle."
Wheeler and Taylor, Spacetime Physics (1966).