Orre, Almia & Relation to Sinnoh

I actually did some research on this during the first lockdown (I needed to keep my curious mind busy in those difficult times lol). I replayed & datamined some games and I found that not everything you find on the net is correct regarding Pokemon geography.

But tbh we'll probably never get an answer because it's not like the designers who came up with Almia will ever do an AMA or an interview to give us insider info on a 10+ year old game.

Sadly you're probably right. Maybe we could reach out in some way to Director Katsuyoshi Irie or Producers Hiroyuki Jinnai and Hiroaki Tsuru since they seem to be involved in all forms of spin off stuff and they may be still willing to answer on these forgotten regions, but I wouldn't count on it.

I thought Almia was part of the Sinnoh landmass, but that's only based on its shape and the fact that a lot of Gen 4 Pokemon live in Almia.

I've always thought the same, however I'm not sure anymore after considering that this would make Haruba Desert (in Almia) south of Sinnoh and north of Johto (since it's confirmed that Sinnoh is north of Johto). Something like this:

This doesn't seem right climate-wise. Since we are in the Northern Hemisphere, Sinnoh is colder than Kanto and Johto, which are temperate. It doesn't make sense to have a hot desert in between them. Real Japan has mountains there, and this is also suggested by official games and anime maps of Kanto/Johto.
Also, two other places are confirmed to be between Sinnoh and Kanto/Johto:

• Sinjoh Ruins, which is a very cold place north of Lake of Rage.
• If you consider the anime to be canon, north of Pewter City is the Kingdom of Rota, a medieval-like state that it's more like Germany than a desertic area.

So I don't really know. Sinnoh is never mentioned in Shadows on Almia (I searched "sinnoh" after extracting the text from the ROM), not even in the Manaphy event (while in the same event in Fiore it's said that Sinnoh is "far away").
Weird for a place that is supposed to be near it, there would be npcs talking about how they came from Sinnoh etc.

If we factor in that Almia is potentially the tail end of Sinnoh and potentially connected to Orre through the desert, then that means they're within walking distance but a ferry ride perhaps away from Sinnoh.

I think the desert connection theory between Almia and Orre is pretty weak. Other than the desert, they have nothing in common.
I actually concluded that Orre must be south, near Hoenn, because:

• In XD, it's said that since the majority of Orre is a wasteland, it doesn't have wild Pokémon, so the S.S. Libra is used to ship Pokemon to Orre from other regions. Now, the majority of the mons we find there are from Gen3 and the other are from Gen2 and Gen1. It's very likely that Pkmn are regularly imported in Orre via boats from important ports nearby (Slateport and Lylicove) and/or from Vermilion/Olivine (remember that Hoenn is south-west of Johto: confirmed by an anime map, real world Japan, and the Embedded Tower in southwestern Johto being built by Hoenn people (in HGSS). Plus, in Emerald you can reach one of the Sevii Islands by boats, which are south of Kanto).
• The abandoned ship in Hoenn was named SS Cactus before. It's south of Hoenn, near Slateport. I think you can see where I'm going here.

• Kalos is supposedly based on France and Galar on the United Kingdom (personally it looks more like England/potentially Wales to me)...so those could be feasibly reached by train and be close together.

We have no evidence of a Galar-Kalos connection, but I can tell you this.
Couriway Town, in eastern Kalos, has got a train station that is said to "bring trainers from a long way away here". It's implied by various dialogues that when Sycamore first went from Sinnoh to Kalos he arrived just there (leaving the hidden message that you can find in the station).
If we consider that Sinnoh is north of Johto, he must have took a train from some station in southern Sinnoh not visible in the games, then probably passing through Goldenrod station (where the magnet train for Kanto also arrives) and going west for Couriway.
Kalos being west of Johto would make perfect sense. It's got lots of Gen1/Gen2 pokemons, much more than other regions. Everything between France and Japan probably doesn't exist in the pokemon world, or it's shrinked, otherwise Sycamore would've taken a plane instead (they exist in both the games and the anime).

We could place Galar north of Kalos, like you said, and that would mean it's directly west of Sinnoh. That would make sense since the Regi temples in the Crown Tundra are similar to the one in Snowpoint. We should rotate Galar 180° so:

• being in Northern Hemisphere, the more north you go the colder it gets
• the Crown Tundra is north, around the same latitude of Snowpoint (also given the snowy weather)
• the Isle of Armor is not in the way between Galar and Sinnoh, and is where Ireland is in the real world (also it's got the same shape)
• Crown Tundra = Scotland in the north, Wyndon = London in the south, makes sense.

• Alola, if we go that it is based on Hawaii is likely somewhere southwest of Unova.

Likely, and it's also somewhat near Kanto/Johto, since Malie city is said to be influenced by those cultures. Also lots of Alolan forms are mons seen in Kanto.

• Oblivia Islands, south of Fiore and Almia (again according to its page).

I can confirm this as I've read it somewhere in the game, but I don't think they mean Oblivia is directly south of them. More like "Italy is south of England" kind of thing.
I would actually put Oblivia near Alola as they have strikingly similar architectures. Check this out: Compairson of Oblivia and Alola's 1st islands | Pokémon Sun and Moon
Sadly I couldn't datamine the Oblivia game as they used a different format than the older games and there are no tools that can easily extract the text from the ROM.