I've said it before and there are tons of options for it.
Your example options are, in fact, what I'd cite in saying that Greyhawk doesn't lend itself to new player options. They're all the sort of thing a writer comes up with when faced with "Okay, it's been assigned that we're doing this setting, how do we fill these pages assigned to player options?" They are not the sort of thing that are suggested when a writer goes "What should players be able to play in this setting that the game does not already adequately support?" And certainly not from a product planner's "What player options highly characteristic of this setting are compelling enough they would sell copies of this setting book to people not planning on playing in this setting, just so they could play that kind of character?"
I mean, you tell me you want to play a Knight of the Hart (or any of the other Greyhawk orders) in D&D 5, and as a starting point I can already direct you the Noble background (Knight variant), the Knight of the Order background, the Banneret [Purple Dragon Knight] fighter subclass, the Cavalier fighter subclass, the Samurai fighter subclass, the Oath of the Crown paladin subclass, and now the Krynn UA material. I'm sure that a Knight of the Hart
can be differentiated from them, but it would sure look like differentiation for the sake of filling word count.
Agent of the Circle of Eight background? Looks like you're straining to differentiate from the existing "Faction Agent" background.
Multiple Horned Society subclasses might make sense, if I thought WotC was actually burning to do Player's Option: Evil Characters material, particularly for the same-year release as the 50th Anniversary edition.
Cleric domains are easy to proliferate, but that holds for any setting with multiple gods.
And the game doesn't differentiate warlocks who have demons patrons from warlocks who have devil patrons, but "Warlock tied to Iuz" is different enough that the Fiend patron doesn't already cover it?