So, for non-LARPers in the crowd, I want to explain culture packets in a few words, so that the rest of this post will make a damn bit of sense. A culture packet is a document describing one of a game's cultures, organizations, or races, with details on a variety of topics: government, land, history, cuisine, faith and/or superstitions, traditions (birth, death, and marriage at minimum), attitudes toward outsiders, clothing... the list goes on. The list goes on so far, in fact, that actual completeness is no more a goal here than in, say, an introductory text on American culture - or the Silmarillion as a text on elvish culture. Instead, the text presents salient details that might influence a player's costuming choices, details of roleplay, and otherwise establishes a common frame of reference and baseline for the culture.
Because the text has to have an end, a culture packet presents a stereotype, with writing quality increasing the nuance of the stereotype. A recent article in Temporary Hit Points put forth one player's realization that playing the stereotypes of a setting really is better than playing the contrarian outlier. There are a lot of good reasons to learn and live by a culture packet in a LARP, while in a tabletop game most players would look at a multi-page race or culture briefing as more than they wanted to read. In a tabletop game, it's reasonably likely that the elf in the party is the only elf in the party, so nothing is lost if other elves do not share common idioms with her. Even if there are two elves in the party, they develop their idioms together, because players in a tabletop game hear 100% of the dialogue that is relevant to the game, while characters in a LARP really do spend time away from one another, and there may well be multiple full teams from the same culture. Verisimilitude demands that these characters have some sense of their shared origins.
Given that core goal, there are two valid but mutually exclusive directions a LARP committee can go with writing culture packets. One model, used by Shattered Isles, King's Gate, and Eclipse, is to solicit volunteers from the playerbase to write the culture packets, preferably (though in SI, not absolutely) volunteers who plan to play PCs from that culture. This model has a number of strong points.
Once a group of players has written the culture packet, they have a deep level of investment in that culture and their characters. The ideal volunteers for this kind of work are people who were going to do some serious legwork researching the culture regardless of whether or not they're writing the culture packets, so in a sense they're compiling that text for everyone else. (This is not meant to impugn the creativity of those who write such culture packets.)
Plot can get a sense of what their most invested players care about most strongly in their cultures, judging by the focus of the text.
Some would argue that adding more creative voices to the melting pot is a benefit in itself.
It's a lot of writing that the Plot committee doesn't have to do during the same span of time that the Plot committee is also planning their events.
This method also has some observed weak points.
The task has an endpoint, and (in goal if not in practice) that endpoint comes before the first event. The downside to this is that players who join the campaign later, or players who start a second character, do not share that sense of authorship, and elements of the culture that are important to them may see much less emphasis in text and in actual play. This becomes particularly thorny when the culture includes a political or religious organization with a strong emphasis on orthodoxy.
Because the writers do not have privileged access to plot secrets, the culture packet has little in the way of secret information known only to that culture.
Because the writers do not have privileged access to plot's development of the campaign's timeline, writing the culture's history of interactions with other cultures can be very difficult. From what I've seen there a couple of different ways to handle this.
The committee can write the timeline in detail and hand it to the culture writer. This reduces the amount of work that the committee isn't doing, of course.
The culture writer can include a number of conflicts that are not mentioned in the culture packet of the party of the second part. (This is not recommended, as it subverts the consensual reality that LARPs attempt to construct.)
The committee can mention one or two significant conflicts, and otherwise have a world with vast periods of either peace or lost history.
Because the Plot committee may be developing and clarifying their vision of the world even up to a relatively late stage in pre-production, farming out writing makes it more difficult to enact changes that become necessary for inter-cultural consistency.
In all cases, a committee member needs to be in the loop on what the culture writers are doing and providing them with information that it would not be safe for the culture writers to invent. This can come to represent the same workload as if the committee member had written it from the start.
One other factor that did not enter into our consideration, but can be an issue in player-written culture packets, is maintaining a balance of powers between the cultures. Much as a plot committee would typically not let a player get away with describing a force of assassins at his beck and call in his character history, the committee also would not want to see one culture given superiority in arms, magic, the arts, or other fields without their direction. This is not the same as every culture believing itself to be the most refined in manners and customs (as every culture of KG believed, providing a constant source of needling one another) - I'm speaking of granting objective superiority, as everyone wants to their culture to be awesome.
The other practical model, obviously, is for the committee to keep the writing in-house. The Wildlands campaign has always used this model, and Dust to Dust is taking a page from its playbook here. From some perspectives, this model seems to indicate a lack of trust in the writing skills of the playerbase, but as this was not remotely a factor in the decision of the DtD committee, I want to lay out the strengths (and weaknesses) of this model.
Secrets. Some secrets are known to one culture but not another, and some secrets have the two or three hints that reveal them divided between an equal number of cultures. Some are hidden in plain sight. For a culture packet writer to put the information into the text, including obfuscations, they would need to understand the secret in its entirety. If a committee member took the writer's text and added in secrets later, the importance of those changes is exaggerated, and you wind up with some parts of the text that matter and some parts that don't. An integrated whole requires that all writers be on the same side of the Plot fence and free to receive information.
Plot's own vision of the world matures by leaps and bounds in the hothouse of culture writing. It is more important for Plot to have a deep and nuanced grasp of the world's cultures than for a subset of the players interested in a culture to have a similar grasp, in the same way that it is incumbent upon Plot more than upon those same players to provide exposition of that culture to the playerbase as a whole. We have had to write a huge number of NPCs, vignettes, and synopses in the process of writing these culture packets, and the world has been enriched thereby. In my experience, no one is comfortable playing in someone else's sandbox, making it all the more important for these story segments to come from Plot. Not, then, a quality judgment, but an optimization of utility.
Integrating even the non-secretive inter-cultural connections - for example, varying versions of legends and accounts of historical events - is easier if all involved writers are permitted to communicate freely on the matter. Sharing culture packets with people not playing a character of that culture is a breach of sportsmanship.
The committee doesn't have to wait to see if all of the game's cultures will have representation among the playerbase. Even games with a smaller number of cultures than Dust to Dust have had some cultures unrepresented for many seasons of the campaign, or no volunteers for culture packet writing. This leads to either a player from another culture writing the packet, or a committee member finally giving in and writing the culture packet some number of seasons into the campaign.
As the counterpoint to an increased number of creative voices in the melting pot: limiting the number of creative voices to "more than one but fewer than all" improves the tonal clarity and reduces the complexity of communication. To put that another way, a committee of twenty may have so many conflicting voices as to become cacophony rather than harmony.
We have had players submit draft character histories prior to the publication of the culture packets. Some of the things they have sought to include in their character histories have been significant enough to appear in the culture packet. By comparison, a PC-side culture writer would not have access to the uncensored versions of character histories. (This is a specific case of the first point on this list.)
But really, everything after the first reason is incidental.
As to weaknesses:
It is a lot of work. A really lot of work. I can't really overstate how much work it is. The level of preparatory work I've had to do for events in the past, even events that I'm running, has been trivial compared to the labor that has gone into culture packets.
Games using the other model have historically not had culture packets available at the start of play, so players have had no choice but to send in character histories prior to their publication. Because we do plan to release culture packets prior to the start of play, many players have expressed the feeling that they cannot write character histories without culture packets. This is an issue of perception, understandable but unfortunate, despite the fact that we have encouraged draft versions of character histories.
Because of the level of detail that the committee is putting into the culture packets, players have felt unusually reluctant to assume or make up even very basic details, such as the names of the hamlets where they were born. They also wonder about details it has never occurred to us to address - but then, this is just as likely to be a problem with any limited number of writers preparing text for a non-limited number of readers.
Even from early stages, we preferred the strengths and weaknesses of the second option to the first, but this decision was not made as a criticism of any other game's decisions. This is simply what was right for the game we are running.