There are cases where a challenge asks for some specific inputs/outputs, for instance an array where each element is a tuple or something similar.
This might be perfectly fine and clear in some languages, while it might be impossible in others. Therefore, people start asking for clarifications asking how to treat the input in such languages.
The question is thus clear and concise for some languages, while it is unclear for others.
Relevant:
The situation described sounds like one in which the question needs to spend a few days in the sandbox. Vote to close and comment with a pointer to the sandbox.
I don't think your example fits the question you are asking very well. If a challenge only asks programs to take in a 3D matrix, then there is no unclarity.
Languages without a 3D matrix cannot compete which is bad, but not unclear (although the challenge as a whole can still be good in these scenarios). And we shouldn't close vote questions just for being bad.
A more fitting example would be a challenge that discusses keywords. What is and isn't a keyword is abundantly clear in some languages and unclear in others. In this case the question is unclear and should be closed and sandboxed depending on how unclear it is. If we allow challenges to stay open because they are clear in one language we will never get the issue fixed.
