I think the Letters thing is overly complicated, the problem with more bands is it results in the difficulty of one student getting 63% and getting a B-, one getting 64% and getting a B, one getting 67% and getting a B+ etc. Is the difference in quality between the 63 and 64% assignment actually noticeable? How much better is the 67% piece than the 64%, enough to warrant a different grade allocation? It's difficult enough as it is and is why my university changed from a percentage based feedback system to a "high, mid or low" for each classification, which obviously wasn't official on our graduation classification, but was done in our feedback. The problem with percentages is when someone gets say 64% and someone else gets 65%, the student who got 64% desperately asks how they could have gotten 1% more. Our lecturers honestly said at times they couldn't answer that, because it''s just 1%, it might have been how they felt on the day, it might have that they thought "Yeah this is around mid 2:1, I'll give this one 64%", then a few hours later, "hmm another mid 2:1 assignment, eh I'll say 65%". They may been practically identical, but when you're marking a essay based assignments it's practically impossible to have some sort of 1% scale where you can define what a 63% is, a 64%, a 56%, a 73% etc. You might results like that for tests with certain answers, like in maths, but not with essays.
Personally I feel there should be something to show the difference between a "low" and "high" score within a classification. So say 60-65 is the lower half, 66-70 is the upper half, this could be reflected by 2:1-Lower and 2:1-Upper, for instance. I mean I was given a first but I was lucky, they rounded it up from 69% because I got a first in my diss and most of my 3rd year modules. I openly admit someone with say 78%, must have constantly produced higher quality work than I did. Something should reflect that and I'd have no qualms about my degree classification saying "1st class-Lower".