I am still investigating this. I'm going to return to inhabited space to pick up a ship with longer legs as I'm covering a lot of Kylies (and I'll need to do some pamphleteering for everyone's favourite princess while I'm there.)
I was completely wrong to think that the NN-N bit is relative to the galactic centre, it's relative to the sectors.
Also I hadn't appreciated just how many stars are in one of the sectors; very very roughly I make it about 10 million stars per (proc-named) sector. They are over a Kylie on each side after all.
I think that maybe the systems with T Tauris and Herbig stars are being treated as "the mass they'll have once they've finished accreting", so that it goes:
Young system where primary is still a T Tauri - the ZAMS mass of the primary is
higher than its current mass.
System where primary is on the main sequence - the ZAMS mass of the primary is
about the same or a little higher than its current mass.
Old, evolved system where primary is a stellar remnant - the ZAMS mass of the primary is
higher than its current mass.
This would explain why low mass main-sequence stars (GKM) fit neatly onto the letter code scale but young and evolved stars are "higher" on the letter code scale than would otherwise be expected.
I've found interesting patterns in the distribution of stars within a group (e.g. DOOBRY AA-A D0, D1, D2, D3, (...), but that's another story and confusing me even more.
(edited to add) Elite was worth every penny for giving me this chance for fun.
