To add to Goa's post, there's also the question of which lore is even considered canon. With reference to Star Wars, some people may figure that the Clone Wars CGI series, save for possibly a few events or ships, is entirely non-canon due to irregularities in the show and as the people in question may find it to be drivel.

Then there's the question of what to do in conflicting or seemingly conflicting statistics. For example, there are some that consider the Imperial Star Destroyer to be a fairly slow ship, comparable to say the Dreadnaught which has a reputation for being a turtle. Others think that, for it's tonnage, that it has a relatively good speed, based on it keeping up with a CR90 corvette in the opening scene of A New Hope.

To go even further, there is the question of what defines "slightly wrong", which can range from anything from a purely visual change, such as scaling, to actual gameplay changing stats. To bring this to it's extreme, or some would argue ultimate conclusion, there lies the question if a given characteristic, say the speed of strikecraft, is definitely defined in canon, should be implemented as is if it is known that doing so would result in bad game mechanics. In Sins in particular, strikecraft mechanics are widely known to be wonky at best. In vanilla Sins, this is very obvious where on a competitive level massed fighters are virtually useless compared to massed bombers, as they cannot hope to dish out the damage bombers can, nor can they effectively intercept bombers before the deadly alpha strike. Going back to Star Wars, strikecraft have similar speeds, and depending on what source you use, that the famous X-Wing and Y-Wing have comparable top speeds and accelerations that are not that vastly different. However, implementing those speeds as is into Sins would result in pretty awful strikecraft mechanics, and as a result it would be more appropriate to make the Y-Wing slower than the X-Wing, for example around 44% slower.

All in all, I would argue that sticking to a mix between lore and gameplay is important, however sticking to lore as much as possible is preferred. Further, if a major gameplay change is needed, such as adding ships to a faction in order to properly fill out it's fleet roster, it is better to take advantage of lore ambiguities, such as creating whole new statistics for a ship that exists purely in name, than to actively subvert lore to the benefit of gameplay.

In all projects I have created and will ever create, the writing is first and foremost the driving element of all other elements within the game. The other four elements - Graphics, Sounds, Gameplay, and Computer AI - hold equal significance but all serve to drive the story in some way. I have never encountered an issue where writing held back design.

That said the writing behind almost all games is laughable at best.

If you're making a mod for sins expecting to be able to have any choice in how 90% of the game works you're in for a bad time.

If you want to make a story-oriented campaign, then yes to all of the above. That is how their stories went about. Of course you can just make a canon-centric mod and ignore the progression of the actual story and work inbetween. Such is the flexibility of being a writer who doesn't glue himself to episodic definitions.

Also for the record I wouldn't be caught dead making a "fan" mod so I naturally avoid a lot of potential goofiness.

Also, asymmetric balancing tends to win out over symmetric balancing. This is one thing where lopsided lore can really help you make interesting gameplay as opposed to the incredibly stale attack-move slugfest of vanilla sins.

addendum - I don't consider sins a 4x. Sins is like a really, really basic RTS that has no combat depth.

/edit

Reading your post again I think maybe I am not clear in my definition between lore and story.

Lore is the technical foundations of writing. So, in Black Sun, lore is what dictates the Anahn's technological progression, strengths and weaknesses, blah blah. Story is how that lore plays out in action. A story-based project on Black Sun is entirely impossible in a game. A lore-oriented project is somewhat possible.

A story project would always be a campaign. It follows the story preset by the involved writing. A lore-oriented, or as I would call it a to-canon project, would pick a point or attempt to cover multiple points (through, let's say, a tech tree) of objects and elements in that story without actually maintaining the story as it plays out.

In the example of the Death Star, you could say in a mod that multiple death stars may be built and later ones were upgraded with appropriate anti-fighter defenses. The story is irrelevant in a mod because a mod doesn't re-enact the story. The decision in writing what elements dictate what is determined by the creator when building the mod. If he chooses to follow the story, he can't follow the story through a mod like so. Sins is not capable of having story-canon elements because it doesn't allow campaigns.

If I have Phantoms, Xy, Vyru, Templar, Anahn in a Black Sun mod, I break story canon because Anahn never fight Templar or Vyru. But I can keep the mod canon-centric by having the Anahn at a huge disadvantage pound for pound to those opponents in a conceived fight by considering lore-wise how the engagements would go (pro-tip - Vyru are like borg x 1,000 in relative power scale of this world, I'd never add them as a playable race for that reason and I certainly wouldn't nerf them just to add them).

So in the case of Star Wars I would refer to the manuals over the movies to determine the capabilities of units.

When I decided to create story-driven projects exclusively I stopped making mods and focused on conceptualizing campaigns or independent games. I don't find any value in butchering writing to fit a mod. I didn't always think like that, but I outgrew the whole concept of multiplayer over the course of a decade.