For the not-Amber game in my head (let's call it Argent) i have been pondering the baseline element of character creation. For light systems (and it should be light) one of the most powerful things you can do is establish the baseline for what a character can do, specifically what they can do before they start training or specializing or pursuing other things. As an example, if I wanted to do a modern game with American protagonists, I would assume everyone can drive and read and has some familiarity with American history and culture. If the game was more specialized (college students or soldiers) I might assume basic computer skills or combat training. Establishing that baseline makes the rest of design much easier since you don't need to come up with a skill system to represent each of those things. Instead you just assume basic competency (whatever that happens to be in the system) and make note of exceptions, so if a character is illiterate, you note that, not that everyone else is literate.
In a multi-world (and multi-genre) game like Argent, this gets a bit more interesting. Certainly, you have a local baseline, "Scion of Argent" or whatnot, and it probably covers all the basics of a courtly fantasy upbringing - swordplay, horsemanship, poetry, chess and so on - but that's not the only case. Some characters grew up on other worlds, or have spent enough time in them to go native, and the baseline for such places is likely quite different. A child who has grown up on the neon-lit streets of a cyberpunk city probably knows very little about riding horses and has a very different idea about poetry, but he can also probably drive and operate a computer and a gun.
This is not mechanically hard to represent. The creation of new baselines is a simple enough act of bookkeeping, and even games which don't explicitly call this out mechanically can do it implicitly by simply respecting backgrounds. However, I think it's underutilized for one simple reason - every single one of these new baselines is a powerful bit of worldbuilding.
One of the classic problems in Amber, which I anticipate also being true in Argent, is where the opposition comes from. If you establish a place as the center of the universe and give it maybe one comparable power, you've undercut some of the awesomeness of supporting every reality. In many games, powerful groups or individuals come from the infinite spread of worlds in order to liven up games. The problem is that usually those are born from the GM's mind, and they don't necessarily fit the game as the players envisioned it because they had no way to envision this element.
With that in mind, the GM should look to any new Baselines as the places for this opposition to come from. The laws of drama and metaphysics both favor the idea that places become more important from visitation by scions/princes/whatever, so the trick is to extrapolate a power in shadow that uses that baseline.
A good example of this can be found in a comic I like, Invincible. Among its many features is a world-spanning government organization that deals with serious threats, something akin to S.H.I.E.L.D.. It's role in the setting is interestingly gray, and if a character comes from this sort of setting, that organization might be a great element to bring into the game. They find out about other worlds ("dimensions", as they see it) and their agenda of protection and info/power gathering starts extending on those lines.
The bottom line is that whatever makes that world interesting to the character should also make it interesting to the game as a whole. Sure, it creates the possibility of direct ties to the problem, and that's great, but it doesn't necessitate them. The personal relationship to the problem can take any number of forms.
As a bonus, this can also be turned on its head by GMs who already have ideas they want to bring to bear. All the GM needs to do is write up the baselines for the appropriate world. It won't do anything to necessarily warn them of the coming threat, except perhaps in the broadest of terms, but it puts the idea on the table early, turning it into a point of collaboration, not dictation. And players can always say no, but if the GM wants to sweeten the pot by making his baselines a little bit more appealing, mechanically speaking, well, that's just good fun.
This is also a fuzzy idea, but it's also not unreasonable to tie baselines to powers, especially if they're based on a firm understanding of the underlying concepts. This is equally true if powers are mystical things as it is if they're educational. Computer science is powerful magic if you can apply it to things outside of its normal sphere, for example.
1 - This gets a little more hairy when the new baseline is "better" than the normal one, such as a world of courtly aristocrats AND computers.
2 - Need a better name
3 - It doesn't hurt that Invincible, like most supers, already has the idea of multiple worlds, but the model could work just as well with the NSA.
4 - Curiously, this lens also makes the Merlin chronicles make more sense, if you view it as a means to make ideas from our earth (Computer Science) dangerous to the universe.