If I remember correctly, DLL integration is already planned? So if we arrive at the point where 4UC is already fully DLL integrated, then most of the positives of full integration will have already been achieved.
1. Futureproofing - the 4UC mod is entirely dependent on 1 person, me, to maintain it. It’s a precarious situation that depends on me having enough free time and interest to keep it compatible. Moving it into the main mod makes compatibility the responsibility of the people making changes in the main mod. This is much more tenable than the current situation in the medium term.
I admit that I've only dabbled in very minor personal modding, so I might be wrong here, but doesn't DLL integration mean that 4UC will not need Lua anymore and will thus consist entirely of SQL? Therefore pretty much any of the modders could step in if you retire. Hell, I would probably be willing to do that once I graduate a year from now.
As for balancing, I think you are doing a great job of it yourself. Of course, 4UC doesn't have the Congress, but you said yourself that you listen to the issues put forward in the modmod thread, and even in the Congress, your ideas, or their modifications, go through more often than not. Even if you retire and someone like me, who doesn't 'feel' the balance like some people here do, takes up the maintenance of the mod, there will still be the thread where players can suggest improvements. Still, exposure to the Congress is probably the one advantage that full integration has over just DLL integration, but I think it's overshadowed by the negatives.
AI suddenly being able to utilise the new components once DLL integration has gone through will probably shake up the balance. Thus even if we decide on full integration, there would probably need to be a transitional period where the strength (and potentially nature) of the new components is adjusted, or we would end up with another round of reworking something, realising it's too strong/weak and tilting over in the other direction, like happened not too long ago with Rome, China, France and I think someone else.
... if you really don't want the extra components then it is trivial to simply remove them; much easier than adding them in.
Edit: ok so I'm dumb and didn't even consider the possibility of removal being handled by a new modmod. Now that I do, the first half of this paragraph is pretty much hot air.
I think you're speaking as a modder, not as a regular player. Sure, just deleting the files is trivial compared to writing them. But for most people, it's not a matter of coding versus deleting files, but of clicking three times on the forums to download the mod and following instructions to find the MODS folder to unpack to archive to, versus following the instructions (assuming the download thread will have 4UC removal instructions) to find the MODS folder, the Vox Populi folder and the required files there. A difference of one click. Besides, there is probably a not insignificant amount of players who don't follow the development of VP and just install the new version once in a while and play it. Maybe they don't know about 4UC and will welcome the unexpected addition. But maybe they saw it once and decided that it wasn't for them, and now they'll install the mod, see the things they didn't want in the past and go spam "how do I remove this this mod sucks" on the forums. This will be the new "why force GitHub" / "why no Steam Workshop" / "mod not user-friendly" debate, except this time the mod will appear to have actively turned away people. We could lose players that way and while VP isn't exactly a business that cares about the amount of consumers, I think it would be a shame to turn away people who could potentially like the mod by forcing optional features onto them.
The experience of seeing that civ selection screen with the the extra boxes is, for a long-time civ V player, an amazing moment. It's the first thing you will see, and it will tell you that VP means business.
I agree, but amazement could serve to turn people away. My first two VP games didn't have any of the modmods and while I have since discovered them and never looked back, I may have decided against even trying VP if it had them included. From other 4X and grand strategy games I've grown very sceptical of cool-looking features that turn out to be just power creep, and while some of the modmods I enjoy using are pretty much that (More Wonders and Unique City-States come to mind, they tilt the balance very strongly in favour of the human), I would have probably turned away VP if it came automatically bundled with them. 4UC is probably the least concerning one, true, but I feel that full integration would be opening a Pandora's box of further integrations, as evidenced by the calls in the 4UC thread for integrating new civs. Extra civs would also be a red flag for past me who just heard about VP.
Of course, the reverse is probably also true, that there are people who look at VP and think "that's nice, but a bit more extra content would be better." It's hard to think of a solution to one problem that doesn't exacerbate the other, but perhaps we could have a stickied thread in the main forum with a list of recommended mods or something (or at least update the part of the FAQ thread that has it, the OP was last edited in 2019).
Overall, I'm not personally against integration, but I also don't want it to be forced on those who are against it. In a perfect world 4UC would be a toggle on the installer, but the devs have repeatedly pointed out the difficulty of maintaining multiple versions of the mod. To me, DLL-only integration seems like the most prudent option, because once we have that, proceeding to full integration would be alienating a portion of the playerbase for the purpose of saving the other portion a few clicks and achieving a slightly better balance some time in the future. I think we should first go for DLL integration, then iron out all the remaining kinks in the modmod and then give it at least a year before deciding on full integration. Some of the sceptics of 4UC could change their mind once the modmod is entirely within VP's code base and thus doesn't introduce any bugs or crashes. Without such preliminary steps I would probably vote against full integration.