|The Normandy Invasions. You may be more familiar with the tiny dot marked "Omaha".|
1) Size and Detail
A lot of shallow sci-fi or fantasy worlds are tiny both in terms of their area and their diversity. While there are certainly plenty of examples (like 99% of the planets in Star Wars), the one that stands out the most to me is the planet Ansion in the Star Wars novel "The Approaching Storm". Ansion was crudely divided into two areas: the "city" (as in, one city) and the "plains". The conflict that the protagonists were there to resolve was between the city-dwellers and the plains-dwellers. This resulted in the protagonists traveling the planet on animal-back to talk to local leaders and unite the clans so that they could hold negotiation meetings with the planetary government housed in the city. This was the entire planet: One city, and a lot of plains, and the plains are small enough to be traversed by an animal within a reasonably short period of time.
Ansion illustrates a lot of problems, but it also illustrates how those problems could potentially solve. Everything about Ansion is ridiculous (its small size, its two groups, the ease with which the protagonists can cross it) for story reasons. I don't mean that it makes sense in-story, but everything about the planet is designed as a place for the story to happen. In essence, a planet is a self-contained unit that holds enough stuff for the story to take place - nothing more, nothing less. So how would we get around this? This is where the idea of focus comes in: the longer you spend in a given area, the more things you're going to have to map out, and thus the larger it's going to feel.
Armageddon, which was the subject of a lengthy campaign in real life, with 40k players all over the globe acting out moderated battles that contributed to a larger war. While Armageddon was still basically a limited planet (there was a jungle area and a wasteland area), the scale of the battles was greatly increased compared to "there's one fight here and then it's over". The length of time spent on the campaign made it actually feel like a planet was being fought over, and theoretically the amount of different areas would suggest some sort of geographic diversity (rather than the same four spots being fought over again and again).
This leads into a key way to avoid issues of scale: leave the rest of the area open. The need for protagonist-centered importance often leads to a small group of misfits saving the entire world single-handedly, but what this does is negate the contributions of everyone else in the world they're trying to save. Leaving hints or evidence of a world outside what the audience sees works better than definitively saying "no, only the protagonists are capable of doing everything, this is the only event occurring on this entire planet". However, this can lead into our next issue:
2) Room to Grow
Sometimes authors do try to avert the whole "things outside the audience's vision" concept. Sometimes they try to establish that there is, in fact, a larger force at work. Sometimes this works, and helps the audience understand that there's a lot more at stake than just the protagonists' success. Other times, it doesn't. The thing about solid numbers and figures is that they leave nothing to the imagination: there's this many, and that's it. Even if it's a big number, it's still a limited one.
Let's look at Ansion again. Would it have been so bad if the tiny segment the reader saw wasn't meant to be the whole planet? Probably. The issue that I have with it is that it takes this incredibly small area, with one city and a few miles of plains, and declares it to be the entire planet. This is inaccurate. However, if they'd gone the other way ("this is just one city and its surrounding area, but it's important because it's a starport") then it might have been more excusable - the whole planet is technically still there, we just don't need to see it. Here's the tug-of-war of importance: one's plausible, but humble, and the other is implausible, but important-sounding. Obi-Wan and Anakin help to save an entire planet! Who cares if that planet is so small it's barely a county? It's a planet!
three million Clone Troopers in the entire Republic army - despite the fact that the Republic constitutes a million worlds, and there's "quintillions" of Battle Droids to oppose them. Even an attempt to fix this fell short - the new number is "a million droids per factory per year", but the idea that there's only a few factories in the entire galaxy seems a bit bizarre (the fact that "Odds" was written by Karen Traviss, who helped provide the initial "3 million troopers" number is also a bit suspect). In a situation like this, the author must weigh the advantages and disadvantages of stating a solid number. The advantages are that the audience has a clearer image of what's at stake and the world thus feels more "solid" or "concrete". The disadvantages are that the world might feel more limited or constrained - especially if the number is inaccurate or unfeasible to begin with.
One thing I liked about the older editions of Warhammer 40,000 was that it had a lot of room to grow - there were countless worlds, countless regiments, countless everything. This had a very specific role in the meta-sense: everything could be "canon", no matter what. If you painted your own regiment of Imperial Guard, then that's fine - they're a regiment that's just as plausible as any of the "established" ones. If you made a chapter of Space Marines, then that's that - they're your chapter, and they exist alongside all the other ones. If a writer or player wanted to make some new characters or a new battle or even a new campaign, then that was okay. Even new species were occasionally acceptable, as shown by the various minor alien races that populate the galaxy. Warhammer 40,000 was a galaxy with very few limits, and this meant that anyone who wanted to contribute to the collective universe was basically free to do so.
However, at some point 40k started trying to get more into limiting the universe. For example, they started introducing special characters like Ursarkar Creed or Marneus Calgar - specific individuals who were meant to be used in a general sense. You're having a battle with your customized unit of soldiers and for gameplay reasons you bring along Commander Dante rather than making your own unit leader. This negates a lot of the personalization - it's no longer "your unit", because the gameplay rules changed the developing story dynamic. If one of those characters dies in your battle, it's not real - it's just a game. No way can they kill a major character off because he got caught up in some random battle. The whole setup now makes no sense, because it's less and less about your story as part of a larger narrative and more about "here are things that everyone who plays this can identify with".
3) Visualizing Scale
One of the underlying issues of scale is that it's simply hard to imagine or understand. There's a certain point where numbers just become "a lot" - what's the difference between one hundred people, one thousand people, and one million people? Sure, you can identify the number difference, but can you really visualize those differences? It's easy to imagine a battle for a village or a bridge or a beach, but it's hard to get a sense of size when an area is wide or open or sprawling. There's just so much "empty space" - fields, forests, city blocks - that it's much more visually compelling to find an identifiable point like a church or a chokepoint like a pass.
This is probably the reason for the aforementioned issue of planet size, or battle size, or any other "this should be bigger, but it's not" scenario. Fiction is primarily character-centric, therefore characters should handle the heavy lifting in a given situation. However, the amount of ground implied in a planet or even a large island is too much for a few characters to handle. Authors therefore try to find some way to make it about a single geographic location, so that the influence of the characters can be clearly established. It's true that small groups generally deal with smaller locations (you can hardly expect a group of five to handle an entire city), but the idea that "bigger locations = more important" results in errors in judgment.
While so far I've praised 40k in terms of scale, this point leads to one important issue: the Space Marines. The space marines canonically exist as 1,000 chapters that are each 1,000 strong - leading to a grand total of one million space marines across the entire galaxy at any given time. Unlike Star Wars, this is at least justified by space marines being super-powerful special forces, rather than cannon fodder. They can be used for special operations, assassinations, tactical strikes, or whatever other precise task can use a hard-hitting strike force. The imperial guard makes up the bulk of the Imperium's forces - the space marines exist as specialty troops and exterminators.
Naturally, this isn't always the case. There are plenty of examples of authors, players, and the game rules themselves treating Space Marines like "regular soldiers, but better". Space marines die in huge numbers because they're treated like normal infantry, rather than being deployed for the exact operations that they're best suited for. In "Steel My Soldiers' Hearts", an autobiography by Colonel David Hackworth, there is a scenario described where a company commander used a skilled sniper with 40+ kills as a standard rifleman. The sniper was killed, as many riflemen are. This was avoidable because his skillset, if used correctly, would have kept him out of harm's way while still being able to do damage to the enemy. A similar problem arises for the Space Marines - they're treated like normal soldiers when there's no reason for them to be, which results in them being killed in numbers far greater than necessary or plausible.
This is a scale issue because there's no way for the Space Marines to have the kind of numbers necessary for this. In one novel, the Salamander chapter sets itself up as "good guys" because they help defend a refugee convoy - a convoy containing millions of people. How did they defend that whole line? Yes, they've got better armor and weapons and reflexes than a normal soldier, but they're still individuals. This is like having a super-prototype mech and using it to guard a backwater base because it seems like the right thing to do. Not only is it wasting resources, but it's also insufficient to do the job no matter how advanced it is. The numbers are too small - you can't make up for that with "being tougher than everyone else".
Compare this example to most other works of fiction. The Salamanders were made to help those civilians not because it was sensible or effective, but because they were the protagonists. A similar issue arose in the Halo book "Contact: Harvest", where a single platoon of militia is basically used to guard an entire planet because of the need to have the protagonist, Sergeant Johnson, do anything of importance. There's no sense of how huge the planet is - it's the equivalent of a small town with five cops fighting off 20 guys, but the need for "planets = important" means that now it's a small town battle over the fate of a whole planet!!
So let's tie these three things together:
Size and Detail: It's better to develop a location as a combined aggregate of sub-locations, because it makes the world seem more diverse while maintaining a "smaller" scale.
Room to Grow: Leave some areas unexplored and some numbers not given, because it's easier to build on more parts to a whole when less things are forbidden or ruled out. Leave some room for creativity and influence, rather than stating that an entire setting works exactly one way.
Visualizing Scale: It's hard to picture larger numbers - so don't pretend you're using them. If your show or movie or game is about one squad, then have it be about one squad. Don't artificially inflate the importance of an objective if a more sensible, grounded one will do.
When it comes down to it, here's a solid, simple baseline: people connect with smaller groups. Larger groups are really just made up of smaller groups. Therefore, if you show off a lot of smaller groups, the connected effect of all those small groups will help make it feel like a larger group - because it will be. A company of soldiers is made up 75 to 200 soldiers, or three platoons and a HQ unit. Each platoon is made up of 16-50 soldiers, or 2-4 squads. Once you get down to the squad level, you've reached a unit that the audience can easily understand and connect with. As you build up the audience's knowledge of each squad, they begin to feel more like humans, and thus the audience comes to know the whole unit by knowing all of its components.