That’s the genius of “No Style” in a style-based system. You’re not changing the default style, you’re changing the default formatting for any new text you create – and you DO NOT have to reapply that formatting elsewhere. (The fact that people think you do is a byproduct of how we’ve been trained by all the other software we used.)
To the compiler, it’s all the same, because Scrivener isn’t WYSIWYG. The split between the editor and the compiler is such a fundamental concept to the Scrivener workflow.
I can literally have every single document in my manuscript use a separate formatting baseline, and as long as I don’t attach style information to any of it, once I compile it it’s all uniform and exactly as I want it for my output format. If I’m pulling together my documents from a variety of different sources and tools, I don’t have to spend the time to manually make the formatting consistent. Scrivener will do that for me during the compile phase.
There’s no easy way to achieve that same result in a system that forces everything to have a style. Scrivener is literally making life easier by only making you configure the differences that are going to matter.