In my classroom work, I often see unlabeled answer choices such as you show in your print sample. Since the answer choices are printed on a new line, yes I too see them as displayed text. Since they are unitemized (unlettered/unnumbered) you can follow Braille Formats or you can use a Nemeth Code format.
Here is how I see your braille examples:
(My comments having to do with Braille Formats refers to the original 1997 publication. I do not know if the new guidelines will change any of these directives.)
EXAMPLE 1 follows NC for individually displayed items--each starting on a new line in the primary display cell.
EXAMPLE 2 is the same only it adds the BF directive to insert a blank line before and after each "list." Blank lines are unnecessary when braille margins set off the items as they do in NC displayed layouts. Also, if you were following BF list format, the left margin should be cell 1.
EXAMPLE 3 displays the answer choices as a so-called "run-in list" (BF97 terminology) separating each answer choice with one blank cell. This is acceptable, but note that the runover cell should be maintained in 5.b. (the third answer choice should be in cell 9).
EXAMPLE 4 is like Example 3 only with two blank cells separating run-in answer choices. BF97 tells us to insert *three blank cells between widely spaced apart words or phrases printed on the same line. As far as I know, we insert two blank cells only in the context of columned material, which this is not, or in glossaries when the definition does not start with a capital letter. Regardless, since the choices are clearly distinguishable by their enclosure signs, I see no reason to insert extra spaces.
..drumroll..
I vote for EXAMPLE 1. Not only does it follow a familiar Nemeth format, but because there are not always the same number of answer choices it makes it very clear how many choices there are.