Wilderness as Dungeon - Monster Rooms, Strongholds as Treasure?

Previously in this train of thought:

• Never Show Them The Map (2016): Showing your players the ground-truth hex map ruins the joy of exploration.  Dungeoneering gameplay gets this right by making them map for themselves.
• Applied Wilderness Theory (2017): Trying to build a microsandbox borrowing some elements from dungeon play; thinking about days as like turns, and dividing monsters into beasts, sentients, and scary monsters, with the same ratios that those appear in dungeons.
  
• Further Thoughts (2017): Thinking about stocking using the specials, traps, unguarded treasure, and empty ratios used in dungeon stocking, as well as inverse-jayquaying, creating barriers to otherwise mostly open wilderness movement.
• Revisisted (2019): The realization that the hex is more like a 10' square than like a room.  A repudiation of microsandboxing (for reusable wilderness / campaign play).  Considerations around "rooms" or "biomes" in the wilderness.

BH2 got me thinking about the wilderness game again.  Something I've never really worked out to my satisfaction before with the wilderness-as-dungeon idea is how you handle "monster" rooms, if you're following a dungeon-like stocking procedure.  Putting a special somewhere in a 7-10 hex biome is easy.  But it wouldn't make any sense for a wilderness area to have no monsters - surely there are a few bears in the forest.  So then what differentiates a "monster" room from a regular wilderness area?

Two ideas eventually merged.  One was from Koewn's comment on Simple Domains that they looked a lot like monster lair entries, but really beefed up.  The other was revisiting Beyond the Black Gate's How I Hexcrawl posts while thinking about False Machine's Crypt of the OSR post, about the bloggers of old who no longer post.  BtBG noted:

With wandering monsters, I like to have certain "iconic" monsters for each area (such as werewolves in the Blighted Forest and Ankhegs in the Sunken Hills, etc), rather than a mixed bag, as it gives those areas a more distinct flavor. I would then assign a chance for an encounter based on my perceived density of the local monster population.

I think that's probably a reasonable way to handle a monster-heavy biome: roll once for monster type for each "monster room", and the biome is crawling with them.  I'm considering calling it an infested biome.  It might be goblin-infested or panther-infested or man-infested or wyvern-infested or whatever.  All lairs in that biome are of that type of monster (although contra ACKS, I think at most one lair per hex is a reasonable concession to manageability).  The random encounter frequency is higher than normal, and a strong majority (5/6?  9/10?) of encounters in that area are with creatures of that type.

A man-infested biome might be a borderlands domain, or it might just be full of bandit camps.

It might make sense to have a separate table of monsters which can infest a biome, or which are interesting as signature monsters.

Non-infested biomes / non-monster rooms still have monsters, they're just a grab bag and not present in the same density as in an infested biome.  Roll random encounters and lair chances normally for terrain of that type.

If about 30% of biomes are infested (the same proportion as monster rooms in the dungeon), and each biome borders 4-6 other biomes on average, then most infested biomes will have about one or two infested neighbors.  Border tension!

Something else I've been considering is ruined strongholds as treasure.  Yes, I know, they often get used that way already.  But it seems like something that a wilderness-stocking system should take into account.  It would be interesting to figure the value of a ruined stronghold into the total treasure of an infested biome (and to say that most biomes infested with sentients will have some amount of fortification), and to make taking and holding a stronghold grant XP, much like building one.  But I have no concrete proposals yet.