How we got here

Discrepancies abound in The Shadow Elves between the map and the text. (I also found space for a few more lavatubes, castles, and towers here and there - but they don't count, for this purpose.) As for how the map and text diverged as far as they did:

Assuredly some first-edition of the map underlies all the published material. I note that much of the text is interested in tying itself with The Orcs of Thar. So I have an idea on how that first map started: from the lavatubes in Thar (maybe also Glantri). Logically the outflows from the lava lake above had to come through here. Naralf's Hole and Sethandor were marked, going down; and Fire Dragon Mountain serves to push some of the lava back UP. Jorfyn Depression was mooted, for the lava to flow down to, for which location the map's authors figured Fire Dragon for as good a spot as any.

The Blackstone Hollow "checkpoint" coincides where the 792 BC explorers got to Kol, thence to Gnollistan and the Anvil. The northeast Marches, likewise, tie well with the Yellow Orkia. These look intentional. If so, both the Jorlinrath and Estenreth caverns featured on the first-edition map.

The northeast Marches, more important to the first sketches, didn't make the final cut, except for the vestige on the map.

The authors and editors were less interested in any political links further up, to the surface. Elves of Alfheim notes an egress in Darokin near Alfheim. For that, this map has no room - except in the vast neglected south. Losetrel was found to underlie Akorros' suburb Crowlerd, so the Shadow Elves text has run with that.

New Grunland vacillated between cavern and city, and Porador was a farm town... somewhere. I suspect that the "Day" inset was written first, at a very early stage of the map. Later, Porador became a clan, so left the map. Nominally.

Alfmyr was supposed to underlie Glantri. That means its first edition was in the cavern now "New Grunland". Alfmyr moved out, but didn't quite find a place to move TO, except vaguely "south". I suspect the Warrens - which the writers then rejected for Alfmyr too. Near the last minute, Alfmyr got warrens of its own. These were upstream of the east Sojourner with the dwarves. Meanwhile the Cearlian cavern, which just takes up (strategic) space now, may have been first-edition rustic Porador / New Grunland. After the first draught of the text, Alfmyr's former cavern became New Grunland; after its FINAL draught, the map set Jelden / Larkpoint there.

The Dragon Lake was thought out... for its northern shore. Not its south, though. The Lost Souls' desert was ultima-thule, placed where it is before the rest of the south was done. At that point, I doubt it was even the Lost Souls; it was probably just the sandy southern shore of the Sojourner, opposite the Sands Of Qedain.

If so, the Elmglow was original - maybe also its villages, and Sylaros. When the shoreline was split from the Sojourner network, and the long beach became the long desert, this cut off the Elmglow. Beyond that, the south was scrabbled together, to fill up the space left in the 8mile hex map covering the interior foldout.

Last, but the proverbial Not Least: The City Of The Stars. That this vault is upside down has come about because it was a tie-in with the Hollow World. It straddles the gravitational divide between the twin surfaces. The City map ties in with Jorlinrath, likely named late in the process. I suspect much of the confusion here came out of the City's development, and how to merge this with Alfmyr / New Grunland 'neath Glantri.