I read the article. I enjoyed reading it but I always wonder about the usefulness of proposing or speculating about future game ideas. Although I do really like the notion of a seamless world.
I don't know that the style of the game should change (I refer to the dynamic between upgrades and level design), because that's as inherent to the franchise as the empowerment and exploration themes. Maybe even more so; that's what the exploration theme is built upon. Taking Metroid into a world of squad combat takes it away from what Metroid is; make another game for squad combat, or use it effectively, somehow, within the conventions of the game.
I think stealth could work within the game's design, but I wonder whether it's really worthwhile. The solitude and puzzle-solving aspects of the game feel enough like stealth already. It may be worth trying, but I can't say for sure. I think it's worked in Zelda games (which work on a very similar dynamic), but whether it could be a significant portion of the game (like half), I don't know.
Story elements I rarely even consider. They're a MacGuffin, essentially, except instead of being an element that propels the story forward, they're a story that propels the game mechanic forward. Now Metroid has a more coherent through-line than most Nintendo franchises (in contrast to Zelda), so it may be worth maintaining some kind of unity or consistency, but I don't think the game should be approached from a story perspective. If a story element inspires a game design, fine (the Chozo training idea is interesting, if unlikely and impractical). But the creation of a game should be "What game design do we want?", then, "What's the best story we can make to fit within that design," and only then, "How can we fit this story into the broader, franchise story?" It may connect only loosely, but you'll ultimately have a better game.
I'm not saying the game designers shouldn't consider potentially enriching elements from previous games, but they should not take priority.