It's a really good article and would work well enough at the subsector scale. The bigger you get, though, the more problems any 2D representation of 3D space is going to become. The other problem, if I read it correctly, is that the map would only be valid from one particular hex. I'm pretty sure if redrawing the map from a different system's hex, you'd get a very different looking map. Can't prove it but that's what spins in my head.
Every time I've tried to do any realistic map for the local stars around Sol, I eventually run into what I called the 'flayed fish problem': two stars that are supposed to be really close together end up on opposite sides of the map.
While I was considering the question, I came up with an idea that I think could work - but not for subsectors and sectors as currently defined. You'd have to go with 49 hex 'megahexes' (Anyone member those from Melee and Wizard?) for subsectors, and to show multiple subsectors, every subsector would need to have a different map centred on it to show the 'new' six adjacent subsectors. And even then, you still have a bit of a 'corner problem'. Easier to explain with pictures...that I haven't drawn... In any event, while it would result in close-to-accurate 2D hex star maps, it wouldn't be a Traveller map in any sense. Closer to a 2300AD map (or set of maps) if anything. Of course, for continuity purposes, 2300 is going off a 40-year-old star map with lots of missing systems and recomputed distances, but that's an unrelated issue.
