My understanding of what Harold did was to use topology to keep distances roughly equivalent, via jump lines.
It works on a subsector scale, for the local area, I did it myself, back in the day, independantly from him, based on Real world stellar data (funny thing, we both lived in the Dayton area in the 90s.)
The process I used (not sure what he did, we only gamed together one Traveller session with a mutual friend named John) was to convert gliese data to coordinates via trigonometry. Then convert that to a cartesian grid, then convert that to a hex grid, then draw in the jump lines.
I realized halfway through:
1) Too much work, for the payoff in detail. Way too much work.
2) If I was going to do this again, I'd design my own SFRPG and use a 3-D coordinate system from the beginning, a la SPI's Universe RPG (TSR Bug Hunters, same map data more or less), or Iron Crown's Space Master.
3) Way too much work.
I'm somewhat of a backyard astronomer. What really pisses me off is the lack of respect for real world stars in most sci fi. Even Paramount's recent (a few years ago) Star Trek Star Charts is WAAAY off.
You'd think people would get a clue. It's not hard to get real world sources.
A helpful little program that I use even 10 years later is CherryView. These two guys that are fans of C J Cherryh wrote it up. You can program stars for it yourself. It's ancient, but that and the venerable (and rightly so) Jim V's Traveller mapping programs are all I personally need for any traveller campaign.