I appreciate that closing duplicate questions is an important part of maintaining the integrity of chem.se. I also appreciate all the help the users of this site have provided me. It has made a huge difference in my studies.
Saying that to say, I mean to be constructive when I say that I wonder whether we're too eager to close questions as duplicates.
Yesterday, I had a question closed that asks a question that manifestly differs from its putative duplicate. I asked 'Why do some quantum numbers limit the range of possible values for other quantum numbers?' The duplicate question asks 'What do the quantum numbers mean?' The only thing that these questions have in common is that they pertain to quantum numbers.
Less than an hour ago I asked a question, that was, minutes later, flagged as a duplicate of a question that manifestly differs from the question that I asked. I asked 'why are electron configurations most stable when there are 8 electrons in the outer shell' the duplicate asks 'Can an atom have more than 8 electrons in the valence shell?'
I acknowledge that the answers to these questions may overlap in some respects however, I think that most answers to almost any two questions that share a tag will overlap in some respects.
One of the stated purposes of SE is to create quality answers to questions that people might search for on Google etc. If the sentence used to express one question and the sentence used to express another question, do, in fact, express the same question, then it ought to be a duplicate. Google handles semantics well enough that it often provides results that include key words that mean the same thing in context that the search terms mean. However, if the sentences express different questions, then, not only are they different questions by definition, but their meanings differ. Google will not render links to both answers when a searcher inputs terms that cause it to render a link to one of the two answers.
Additionally, even if the information in the answers overlap appreciably, the fact that the answers express the relevant information as it pertains to ostensibly different questions justifies separate questions: Two answers to different questions, such that those answers appeal to the same set of facts are not the same answers. An answer to a question provides information as it pertains to the question. The facts are already out there on the internet and elsewhere. SE organizes facts as they relate to questions.
In sum, it seems to me that questions are being closed when they have something to do with another question, not when they're duplicates. Closing questions as duplicates for having something in common with another question does not conduce accomplishment of the stated goals of SE and does not concord with the meaning of the phrase 'duplicate question'.