As I have stated elsewhere, I don't think the "strength" of preferences is that useful, if we are basing it on the typical online test.
From these tests (which tend to be 2-answer multiple choice -- "This" or "That" -- comparing paired functions), even if I am scored as a 100% in N (for example), all this means is that in 100% of the questions, I was at least 51% N.
So a 100% N as listed in this thread really just designates an N that ranges anywhere from 51-100%, correct?
I do not find this very useful, at least for the level of precision that these types of discussions usually demand. (And it is also misleading.)