BTW, I've been wondering already for quite a long time about whether 5e/DnDNext would be a good system for Rokugan. I don't think there's any chance to get a Rokugan Campaign Setting book for 5e, but that doesn't bother me: I have enough narrative material already, and I could adapt/create my own "crunch" for Rokugan.
A key question is whether 5e offers enough mechanics to represent clan, family, school and other affiliations as source of training for the characters.
On one hand, such things could be complete add-ons, i.e. in addition to all the stuff that 5e characters get, in a game of Rokugan they might also be required to pick a clan and get additional benefits. But OTOH it might be better if clan choice was exploiting an existing subsystem.
Here's the characters differentiation subsystems offered by 5e, and my notes on them:
a) classes: a Rokugan campaign might use alternative/adapted classes, most importantly the setting needs Samurai, Shugenja, Courtier and Monk. These could be designed from scratch or as adaptation of standard classes, e.g. Fighter, Wizard, Rogue and Monk. There can be more classes than just four, but multiclassing in Rokugan is not common (from a narrative perspective). Special characters such as Ninja and Maho-tsukai might be better served with a special treatment rather than with a core class.
b) races: Rokugan characters are normally always humans, but in some advanced campaigns Nezumi and Naga might be allowed. In that case Tribe and Bloodline respectively become important choices, which could be implemented as subraces.
c) backgrounds: these may just not work properly for Rokugan, since human PCs are supposed to be all bushi (i.e. nobles). However, with some flexibility on the concept of background (in standard D&D a background is basically your job when not adventuring), it might still make sense to have e.g. Artisan (artistic professions only), Sage, and perhaps a few others. But also, a valid alternative would be just to sort out which Traits still make sense for a Rokugani bushi, and then let each PC pick Trait and proficiencies indivually. Languages and tools may not be common proficiencies at all in Rokugan, which might mean that the standard background benefits (1 Trait + 3 Skills + 3 Tools/Languages) might need to be modified.
d) skills: in general, the skills list would probably need a revision, to make sure there are no "dishonorable" skills. The 5e skill list is very short (i.e. the 5e skills are very broad) and this might make it somewhat difficult to introduce new thematic skills, because they'd have to be roughly as useful as the existing ones.
e) subclasses: this could be the simplest and best mechanic to deliver clan-specific benefits to each character. Basically a Crab Samurai and a Lion Samurai could have different features at those levels where they get subclass benefits. There may also easily be different subclasses for different schools.
f) feats: since these are picked individually by each PC, stuff like ancestor benefits, secret lore or special training obtained from a unique master, could easily be delivered in the form of feats.
g) magic items: compared to 3e, 5e is supposed to allow much more freedom in the amount of magic items owned by the PC, and that is definitely a benefit for running a Rokugan campaign where magic items are on average much rarer but also much more dramatic than in classic D&D settings
h) alignment: in 5e the importance of alignment is largely de-emphasized, and that fits Rokugan very well. Obviously, some sort of Honor system should be added to the game, but IMHO this is an easy addition. For instance, I could just take the d20 Rokugan Honor system and slap it on top of a 5e game. I don't see any problem with that.
i) equipment: clearly we need oriental weapons, however I believe that we can manage by using the standard weapons and just change the names and description.
j) spells: this will be simply a matter of sorting through the huge list of standard spells, remove those which are inappropriate for Rokugan, keep what's already usable, adapt what is fairly close to usable, and eventually add new ones.
