There's something ironic about Jim Sterling complaining about the Youtuber-bait content, but hey. He's definitely got a point though, because as a game itself Dreams doesn't have a lot to offer. It's an amazing tool for people who want to create cool experiences and are willing to put the time and effort, without needing to learn proper game coding. A bit like the custom map and mod scene in many PC games.
There is however a huge difference between Dreams and the crazy WADs people make for Doom, the innovative custom maps that spawned stuff like DOTA or tower defense in StarCraft and WarCraft 3, or even stuff like Trials. Those games are brilliant titles on their own: if you don't want to mess with the editor or play crazy shenanigans made by others, there's an incredibly deep and satisfying base game to mess around with. They're games for everybody, in that sense.
Dreams will no doubt attract a lot of media attention for the remakes and the crazy ideas the community makes. The joke ones especially, like the RDR2 simulator. But at the end of the day, it'll always be a fascinating thing to look a YouTube video about or play yourself for a couple minutes, thinking "damn, I didn't think this was possible", but at no point these projects can compete with the myriad of incredible games out there. Not that they should, mind, but it's worth keeping in mind. It's fascinating to create stuff in it and to follow creations, but at the end of the day it's a bit like watching how people make relatively complex graphics and gameplay mechanisms on old Ataris or NES'. It's fun as an idea, but it's really not that exciting to actually put hours and hours into playing those projects.
I still expect a couple fun game modes to spawn from the game that may even spawn their own genres.
