Given how big some buildings are by necesity (the hospital, the school and the mall for instance), mapsize would increase if you put more locations there. Also, it wouldn't make sense since it's modelled after a US city pattern. While in Europe, "districts" aren't really a thing, they are in the US. Residential houses next to shops aren't common (even though it's, well, more convenient). Adding new locations in new area's would make no sense and just complicate things in the long run. Need a shop? Don't bother looking at the shop area, because there's no shop area, but go and find it in one of the other locations.
You're thinking way too complicated. Maps like the Hospital are standalone maps, so you don't need to merge them anywhere and they don't affect the size of any other maps.
You do imagine this like the town would already exist and the devs have to deal with it (Like:
Oh shit, we need a tech shop, but we don't have a shopping district so we need to put it between houses which is unrealistic for the USA). They're the creators, they can plan it right away.
So let's imagine the first 1/4 of the game requires the school, the hospital, the grocery store, comic book store, tech shop, the gym and some casual houses where people live.
So school and hospital are unique maps just like now. All the stores can be in one shopping mall or district, or you can split half of them in one and half of them in another map (district / mall). Then we have the neighbourhood with the gym (or you can put it into one of the other two maps with the shops).
So you see, no problem getting all these elements into 3-4 map parts. And then you just progress the story first in these 4 map parts, and focus on the characters given there. No need to have another 5 districts with another 10 characters out there that you can't interact with yet, that have nothing more than 1 quest to offer.
As the first 1/4 of the game progresses, the 4 parts of maps are being filled with life and the characters all have something to offer in terms of quests, then the game can naturally expand and unfold into new areas, introduce new characters and make previous progressions more complex.
This would be a way more ideal approach that is more streamlined and structured and offers the people more easy-to-figure and further progressed content rather than running around in a huge empty world looking for little snippes of dialouges or intro-quests that lead to nowhere yet.
Also, this brings up the "problem" with polls again. Previous introduced characters will always win in the polls, rather than new characters. By introducing all the characters as soon as possible, you take away that bias.
There's no problem, you make a poll, give people 5 characters to favour and then, when creating let's say the first additional map with new buildings, start with the first 2 characters the community wants to see the most. Then you slowly add the picks 3-5, and this way the new map and its characters come to life. But all the time with content that really matters and offers more than just a simple intro quest.
Of course, you'd only offer 5 characters that are part of the next additional map that they come with so it all makes sense in the context of the game and the current progression. This is a modular approach and would fit best for early access games that are supposed to be enjoyed and funded by a community during development. Polls are also something that devs have control over, it's not that they just need to deal with whatever the community picks and then flip their whole game over for it.
Like I said this just needs to be well thought through from the beginning, but it's easily possible.