I was looking at the comments for Does transforming keyrunes count towards Evolve? and was curious about when questions actually become similar to warrant a close vote. corsiKa's comment and link point to the idea that the question and the possible duplicate (If a permanent turns into a(nother) creature, has it entered the battlefield?) are distinct questions.
This caused me to go back and take a look at some older duplicate closings. I noticed one that was closed but probably could have had a similar argument made to stay open. https://boardgames.stackexchange.com/questions/8020/globals-and-protection-from-that-color asks about both global destroy effects and global -/- counters, but was closed as an exact duplicate of of What does protection from [quality] do?, which asks only about global destroy. This would make them appear to be distinct questions even though the answers will be the same; "destroy" is not the same as adding -/- counters to creatures, but the result is the same when you're talking about color protections.
The questions about evolve are in the same situation. They could be distinct questions and have the same answer; lands are not artifacts, but the interaction with evolve is the same when you animate either.
So my question is: where is the line on whether a question is close enough to another to warrant closing? I feel like this is at least sort of common on B&GC since we deal with rules questions. Small changes in the object or game state may actually change the way rules are applied, and either of my examples could (although they don't) have different answers because the event is slightly different (destroy or -/- counters; land or artifact). It seems important to find this line to be consistent in closing questions as exact duplicates.