I understand what you're saying. My idea was more for flavor's sake, and I get that it would be very difficult to code given the amount of benefit one gets from it.
To be clear, you're referring to the idea about the [Prestige/Glory] faction potentially demanding that the liege declare war on an enemy, right? That's to what Anax was referring when he said "(which would probably be a nightmare to actually implement in any way beyond just being upset if peace lasts too long)."
The reason that's hard & the reason we wouldn't do it are separate matters. Reason it's hard: Script would have to know which CBs were valid, who were plausible targets per CB, and whether the war could reasonably be accomplished based upon relative power estimates. The first of those (CB evaluation) would be pretty ridiculous (although this is just a modding limitation). Primary reason we wouldn't do it: we don't want to make more out of these type of factions than is plausible. Let me explain the latter for the sake of the general audience here.
What I mean by that (plausibility of faction demands for this style of faction):
The notion that a group of all vassals affiliated solely upon the basis of preferring prestigiousness to piety or wealth (in the example case) would actually meet together under the pretext of only that loose/subtle/implicit affiliation and somehow come to a consensus which leads to a kill-or-be-killed ultimatum vs. their liege . . . it's kind of silly, if you think about it. With standard factions (or perhaps some other kind of association which could lead to revolt over a particular issue), it makes more sense, because the faction's purpose (e.g., deposing an anti-pope) is what binds the vassals together, and they don't need to vote or anything parliamentary like that for their purpose to come to fruition.
'Glory' or 'Prestige' isn't a purpose but instead simply a value, the primary one for its faction's "members." Pretend we're vassals in some large medieval realm like the Persian Empire. You and I might both highly value <favorite political issue here>, and if we chatted about it enough over time and came to know enough others that value it in the same way, we might risk losing our heads sending an ultimatum to and potentially marching upon our liege about <favorite political issue here>. However, if we both just happen to value Prestige over Piety and Wealth, that's a really weird pretext for uniting in rebellion on its own. However, value-oriented groupings could be very useful for assigning liege opinion due to liege actions / outcomes of their rule (and then letting alignment along purpose fall where it may).
If you add voting and make essentially a legal institution out of the faction (like a political party, except essentially and perpetually treasonous by definition of monarchy/feudalism), that changes things. I will note that I don't _hate_ the idea of voting about things a value-oriented faction can do; it can be viewed as an abstraction of consensus just as the standard faction system can. You do really need a way around the 'treasonous political party' bit, though. [The HRE had a few, limited legalistic ways around this, although they are the exception.]
As long as these political alignments can produce the inherent instability that plagued so many medieval realms and effectively halted any efforts at centralization, I'm all for a more minimalist but still far-reaching approach.
I think this potential addition (emphasis upon potential) falls more into the category of adding depth to play than inherently promoting instability. Not that I don't want to promote realm instability.
.png){kind=avatar}
