There is one simple reason: the bigger the game, the fewer other games you can install. So why spend time optimising size on disk, if it will just cut into the live service profits of your game?
I don't think that's a reason they use. If anything, I'd expect live servicr games to benefit from the game being smaller, because they want players to not uninstall it when they're feeling finished.
But "why bother" certainly is. There's not that many players who will not buy your game just because it's big. Most games are focused on the initial sale. Optimizing for size is expensive. Why spend thousands of dollars of expensive software dev time while making the game more complex to test for something that won't affect many sales? Especially compared to fixing bugs or adding more noticeable features. Software dev is always a matter of tradeoffs. Unless you're making something like a mars probe, there will always be bugs. Always always always. How many people will complain about a bug (which unexpectedly turned out to be game breaking in some niche case you didn't think about)?
I don't think that's a reason they use. If anything, I'd expect live servicr games to benefit from the game being smaller, because they want players to not uninstall it when they're feeling finished.
The smaller the game, the more likely you are to uninstall when you're done, because it's easier to re-install (and you don't need to free up lots of space). Large size means you'll keep it installed so you don't miss the next content drop. Keeping it installed means you're more likely to play, because you have less choice.
It cuts both ways. If you're low on disk space, you can uninstall 10 small games or just one large game.
Though personally, I've been happy after I threw in a larger SATA SSD and now I can move installs from my faster NVME drives to the slower one when I want more fast space. I generally ignore the games smaller than 5gb when trying to find more space.