Those games are not tactical simulators, they're games designed and balanced around making getting the expensive cool thing a good idea. In Supreme Commander building a thirty storey tall spider mech with a giant laser on it is a good idea, when in reality it'd be an absurd one, killed by a single high altitude bomber rather than need to be pecked to death by dozens.
P.s. You can feel it in a lot of strategy games, closer to the theme- EaW Remake, placing few ISDs betwees SSD and enemy turns large piniata into chainsaw for enemy fleets.
The attrition curve effects(your point B) in video games are weird compared to a reasonable scenario because typically units fight at 100% effectiveness until they moment they die, or close to it, while this is not the case in reality. There's no lucky hit on a magazine blowing a huge chunk out of the super-battleship and disabling a major part of its weaponry, or degraded effectiveness as it's crew gets worn down.
In any theater and tech band there's a point of diminishing returns where building something larger just gets you a bigger, more expensive target. Supertanks, superplanes, superbattleships, they have all encountered the same sorts of conceptual and engineering obstacles. Ships in space won't have air resistance or ground pressure to deal with, but they do still have to contend with escalating forces and thus structure to get the same nimbleness as a smaller craft. You don't need to be slow enough to be hit by long range mass driver fire for a lack of speed to attract deadly amounts of fire.
There are also advantages to being big, but there's naturally going to be a viable sweet spot and unviable extremes. Star wars doesn't care about that though(and never has), because showing that thing A is bigger than thing B, when we understand that thing B is already impressive and dangerous, will intuitively make thing A look even more impressive and dangerous. It's good visual shorthand, but it is not a universal law of engineering.