But that’s exactly what you’re doing with Section Types and Section Layouts, isn’t it?
You tell Scrivener what you want each element in the Binder to do (be a ‘Scene’, be ‘front matter’ etc) by allocating it a Section Type. Once you’ve done that, you never need think about it again in the Editor, no matter how many different compilation formats you’ll have eventually. With your scenario, you’d be constantly going back to the Editor to reallocate the layouts.
Instead, in Compile you simply say 'format all the scenes (Section Type = Scene) like this (Section Layout), format all the front matter (Section Type = Front Matter) with that Layout. When you compile the document, you are given the list of all the Section Types you have defined and all you do is link them to the appropriate Layout. If you don’t like the default layouts (and they cover a lot of the most common scenarios), then you can edit / create your own.
The beauty is that once you’ve defined the Section Types for a Project, you can compile it to any output format you want, simply by allocating a relevant Section Layout to it (either one of the default ones, or creating your own) in compile.
E.g. In Version 2, if you wanted Appendixes which differed from Chapters only in that they had the Word Appendix and a letter in the title, rather than Chapter and a number, you had to try to force that into the Binder structure somehow (Chapters = folders, Appendices = document groups and so on). It was possible, but fiddly, and not really intuitive for new users.
In Version 3, you just allocate the Section Type Appendix to one, and Chapter to another, then choose / modify the appropriate layout in Compile. It’s conceptually much simpler, but it also makes possible a lot of more advanced differentiation if you need it.
But it is a different way of looking at compilation, which is why it causes experienced V2 users a bit of disruption until we get used to it.