You want your big guns available to pound the defenses...
Yep.
...which are on the slow ships.
Why? The slowness of the ship doesn't improve the ship. It's a gameplay annoyance.
You want your fast bombers to arrive early
No, I want my bombers to arrive at the same time as all my other ships, so we overwhelm the enemy. I don't want them to arrive early so the enemy can focus all their ships on my early arrivals and take them out before the rest of my ships get there.
The ship speeds are a nuisance - trying to work around the speeds is not me fighting against my opponents, and it's not me fighting against the environment... it's me fighting against the game rules.
You want your fast little ships to take your troops from the big slow troop carriers to the beach.
I got a better idea... how 'bout I crash my big fast troop carriers into the beach directly?
Why do I want to add extra ships just for couriering troops back and forth between the attacking vessel and the beach?
If they were automatic, and required zero player intervention, then yeah, I can see that as a neat visual effect for added immersion. But otherwise, it sounds like the game would be simulating in detail a boring administrative task.
I'm really not anti-realism. I love realism in games... within reason. But I don't like the realism of tedious tasks to be integrated into games I'm playing, just because that's how other games in the genre do it. There is some place for tedious repetitive tasks in games... but we need to be sure we actually want them when we add them.
Speed then becomes an intrinsic part of your plan of action. The same should be true in a game. It adds another layer of complexity to the game play at a tiny coding cost.
Indeed so. But!
A) I'm not worried about coding cost - I shouldn't add features to a game just because it's easy to do.
B) I'm worried about player annoyance. You don't want your game to die of 1000 paper cuts of minor inconveniences.
C) Complexity is nice... but there's good complexity and bad complexity. Complexity by the layering of more things to manage is 'bad complexity'. "Entities should not be multiplied beyond necessity". Complexity by the dynamic interaction of a small group of simple systems is better, in my opinion. This latter type of 'good complexity' is sometimes also called 'depth' in games.
You may decide to split your fleets into smaller fleets based on speed. Running the risk of your transports getting intercepted and destroyed, but allowing your fast attack boats to get to the enemy quickly and start doing damage.
So if you group your large and small craft into a single uber fleet, you are slow and easily detected, but very powerful and able to defend your self.
Or you could send your big heavy ships of towards a secondary target and send the harder to detect, light, fast ships to the main target. Possibly drawing the defenders away from the primary target.
Everything here is your reactions and decisions as a player to work around the limitation the game artificially constrains you with. Such limitations are good, because yes, they create gameplay by requiring the player to make informed decisions after understanding the rules of the game world - but you have to consider what limitations (and how many) you are adding to the game, and why you are adding them.
Just because the limitations results in player choices doesn't mean it should be added! Almost all limitations will result in player choices - that doesn't mean we should cram them all in a game.
For example, the limitation of fuel I mentioned above, or the limitation of speed under discussion here, the limitation of health, resources, construction speeds, ongoing ship maintenance, available population, attack range, movement limitations (can the ship go within planet atmospheres or must it stay in space? Can it fly through geomagnetic storms?), and so on.
We can pile on more and more interesting and choice-creating limitations on players, layering on complexity just because we can, or we can consider where the real strengths of the game is, the enjoyable fun parts, and figure out how to cut out the unfun parts, and create real depth by adding extra detail around the core parts of the gameplay.
If we just add the same limitations other games have, without thinking them through, we'll just end up creating similar games. What if travel speed wasn't a problem for the player? How will the game change? Will the players focus on defenses - and instead of walls of defenses, will they have to have defenses inside their walls as well, incase their enemies just warp right past their first lines of defense?
What if shipbuilding occurred instantly, instead of waiting three or four turns? Is construction time one of the limitations of your game that you are intentionally adding, or are you doing it 'by default' because other games you've played have done it?
Health and shields on your units? What about making everything 1-hit KO? I'm not suggesting doing that willy-nilly, but I am suggesting thinking about the limitations and resources that you might be accidentally taking for granted in your game's genre. If everything was 1-hit KO, maybe combat revolves around slow moving projectiles (nuclear space, uh, torpedoes? ), and having your ships try to evade the projectiles, and consciously deciding to sacrifice lesser ships (like a game of Chess) by having them act as a shield and be hit by the projectile. Just a random train of thought by usurping commonplace typical genre and asking, 'Why?' and 'What if?'.