[smsm42]

That sounds like, if you use ternary logic (true, false and CU), then the relationship between sizes of N and R is different. Intuitively, that sounds kinda wrong - why sizes of N and R and their relationship would depend on which kind of logic model you're using to reason about them? If you could build a 1-1 mapping between N and R somehow employing your logic system (which is equivalent to saying they have the same size then shouldn't it be possible to build the same without using your special logic? At least intuitively it sounds so, what am I missing?