#No.
No.
Something like this was tried several times and it failed.
In Russian imageboards, there's idea of 'Битардск' ~ '/b/-tardville' - rural village squatted by anon commune - or collectively rented property. On one hand, swapping society you're not comfortable with to one you are comfortable with looks neat. With some amount of planning commune may even seem self-sufficient, in theory.
In practice, these tries usually fail. Turns out stable online society does not translate to stable offline society.
Participation in online communities is voluntary and there exists a selection bias. Some people are way more likely to participate than others and you're going to have surplus of those and deficit of people who are not likely to participate online.
For example, skillsets of people in StackOverflow community are heavily biased towards programming, which also means that they are biased against stuff that is pretty far from programming.
Well, problem of skillsets can be solved. However, there's another problem: personalities.
People usually get a healthy-ish dose of personal interaction offline with different people, and some interaction online with some people. Online societies in this setup are not expected to provide all meaningful interactions and are not tested for their ability to do so.
Even if we speak about imageboards that can occupy majority of user's social interactions, they're not going to translate into offline societies well. Same thing with supply and demand. If you have oversupply of people who are good at discussing things you're likely to lack in department of people who can actually get stuff done. There could be some bonding thing that partially solves the problem by forcing people to do stuff they won't do usually. However, online societies don't provide that.
Drugs, for example, do and sometimes attempts to build a b-tardsk devolve into a drughouse. However, drugs are just a filler for when lifestyle and society does not provide things you need. Cults provide bonding too. Although they are not online, I'd expect cult-based society to outperform online-community based one, provided that both are trying their best to get by. There are cults (A outcompetes B, A competes for the same resources => B is going to have a bad time).
Online society is also not tested for ability to provide stuff offline society is expected to. For example, sex life. Attempts to round up some willing anons and build a community fail spectacularly if there are girls in the community. Competition for sex generates ridiculous amounts of drama. Same thing, although a bit delayed (because Overton window is inert) happens with groups with gays or bi-curious people, and you don't have any realistic way to prevent that.
I've heard of only 2 semi-successfull attempts - 1 became a drughouse but people who remained there were quite OK with that, and it lasted. Another operated as a web studio, but people there knew each other for some time before and that course of actions was planned. It lasted for some time too.
Speaking of larger societies, I suppose it's even less likely. If communications are disrupted, gathering people from the whole country is going to take years and lots of resources and is not guaranteed to pay off, let alone global project. Yes, some centralisation would be good, but not a global one. Especially when somewhere out there is a thing that would gladly nuke that newfound city.
I'd expect Amish to outperform any online-based community by a wide margin and hence I wouldn't expect online-based community to seize power in Future Empire even if there was built one.
