Well done article, really accurate from my experience.
I'm a level designer that just spent over a year on what amounted to about an hour of a shooter's campaign.
My main time sinks:
Iterating on combat. Subtle adjustments to spaces and enemy behaviour (and wave composition) to make encounters fun. There was also plenty of overhauling and removing encounters. Mostly based on feedback, so lots and lots of iteration. There are also game mechanics and enemies that come in late, so your roughly blocked in version of something may need to be changed completely once you get closer to the end of development.
Cleaning up after artists. I don't mean this in any disparaging way, but when art goes into a level it can mess with things like collision, memory usage, AI behaviour and the way things are scripted. It can also inspire you (or your boss) to change something about the level when you see the art. Like maybe a balcony gets added by an artist, and you decide it's a cool spot for snipers. Now the whole encounter changes.
Memory usage. Splitting assets up into sublevels that stream in and out, reducing types of enemies loaded, etc. Especially on consoles.
Fixing bugs.
Helping every department that puts things into levels. VFX, animation (in-engine cutscenes) and audio are the big ones. Especially since nobody knows the level as well as the designer, so you're probably the best person to implement things without introducing bugs.
And all of this stuff takes time. You might spend a week just cleaning up collision on a level, or a day tweaking an encounter to feel just right... then get told by your lead that it's too similar to something in another level and needs something to make it unique.
I actually love the whole process, but the reality of game development is that lots of people that love playing games would hate making them. As a designer anyway, you need to be as good at knowing when to kill an idea as you are at coming up with it in the first place.