It's interesting, and he raises a few very good points, especially about wasted words, misused characterization, purple prose, and telling not showing. Lux's lore really would have been improved by cutting directly to her interacting with the soldier or making the surgeon an actually active character. And he definitely makes the case that there are just some obvious sloppy mistakes that could really use an editor to fix.
That said, I think he does overplay the problems of fantastical description. While he's right that some of the worldbuilding is rather hollow namedrops, but some of it definitely works. The whole Garen memory bit, where it lists a few battle names, for example. I think that one completely worked, it's purpose was to give examples of Garen's battles and his actions during them. The names provide details that feel more real than saying "And then Garen was on another battlefield with a dying guy" because it's Garen's memories and he would know what the battles are. Adding a few of these details do work, maybe not all of them, but some of the complaints are just complaints.
The other one was comparing the walls to being attacked by a Noxian siege engine. It sets up right away that Demacian know what it's like to be attacked by a siege engine from Noxus. Showing that they're either at war or were once at war with them, and that this war is so ubiquitous even the noble girl's mind draws comparisons to them with tragedy. Unless you're an idiot you know big bad earthquake destroys walls. You should also know siege engines destroy walls. The argument that "we don't know what a siege engine would do to a wall, maybe they're weak!" is ignoring obvious context clues for the complaint.