If the song repeats something so much that simple repetition gets stale, I make
variations of earlier ideas and/or
expand upon those ideas, making something new while keeping enough of the core idea that it's recognisably related to what came before.
In a strict sense, the former method tends to be more applicable to stuff where the song just repeats itself, while the latter is great for spots where the song repeats an idea but in a noticeably different way or with more intensity. Making your ideas evolve can be viable even when the song does nothing new, though: it's a kind of artificial layer of progression in the map.
To give an example, I make heavy use of these methods in
☆☆☆☆☆☆*****, where the song repeats the same ideas a lot, with some differences each time. Let's take a look at how I handle the song's main motif in different parts of the map:
- In the intro, I introduce the basic form of the idea I'll be working with.
- At 00:10:489, more instruments come in and the song gets more intense, so I increase my spacing and rhythm density as well as adapt the rhythm of the pattern to fit the new stuff in the song.
- At 00:30:489, certain background instruments get stronger and the drums start doing 1/4 rhythms a bit more often. I change my rhythm and patterning quite a bit to match the change in the song's vibe, but the focus on overlaps is still there and the general movement is still similar to what I had before.
- At 00:40:489, the song overlays the main motif with the guitar from an earlier section, so I combine the ideas I used in that section with the ideas I've been using for the main motif.
- At 01:20:489, the song is pretty similar to 00:30:489, but I'm evolving the map ever so slightly towards what will come next by using overlaps a bit more sparingly and making my 1/4 patterns a bit harsher.
- At 01:30:489, the pitch of the melody goes up a notch and the timbre becomes sharper as well, so I increase my rhythm density and add 1/4 sliders to the mix.
- Finally, at 02:10:489, we reach the climax of the song, with the same higher-pitched melody as before overlaid with what's basically a guitar solo; it's all chaos from this point until the end. The 1/4 sliders are back, I'm using slider shapes I introduced in the calmer part of the guitar solo, the intensity is dialed up on all fronts, and although there's a lot of different things going on all the time, you can still see ideas from the first half of the map used here.
Now, going back to your post for a bit, you say that after mapping one section, you're just now realising that there's another section just like it. Learning ways to structure maps is good and all, but I think
this part is actually the real cause of your problem.
You shouldn't be realising that two sections of the song are the same after you've already mapped one of them, you should identify things like that ahead of time and
plan your map accordingly. It's often easier to expand upon ideas if you've come up with them with that in mind. Besides, in the first place, you can't really design the overarching structure of your map by just reacting to each section of the song as you get to it; you've got to consider the song as a whole.
To illustrate this a bit, I have to confess that I wasn't entirely truthful in the way I worded my explanation of the map I linked. The way it's worded sounds really linear/chronological; that's the way a player would see the map, after all. However, that's not even close to how I approached making it.
The last section was the first one I mapped, because that way I always had a clear picture of what I'd be working towards throughout the map. After that, I mapped the intro to show the other end of the spectrum, then I mapped the beginnings of several other sections to sketch out some ideas, and finally, I filled in the gaps to finish the map.
Compared to just going through the song from start to finish, mapping the song out of order like this gives you a lot more control over the structure of the map and how your concepts evolve over time. I'd say it's straight up the better way to work if you want to end up with the best map you can with your current skill or if you want to practice some overarching concept like structure between sections. The drawbacks are that it's slower and forces you to think a lot more, so if you just want to chill, practice more detail-oriented concepts, or get a map done fast, a linear approach can still be valuable.