Oh well… 
The possibility that one might take a remark such as
and follow up with the question/discussion of “what could be the meaning of [[i32]]”, even though that’s entirely besides the point of the technical remark @CAD97 was trying to make[1], was exactly the motivation for my clarifying message that
Edit: This seems sufficiently off-topic from the original question of whether a generic “T” includes “[T]” that I’ve split the topic.
a remark which exclusively addressed niche cases where the compiler would currently kind-of accept “
[[i32]]” as a type but did never imply there could ever be any value of type[[i32]]↩︎