But merely getting up to 1G is trivial. I think the reason we don't have it today is that while getting "something" up to 1G is trivial, all the other engineering things that it interacts with get to be non-trivial, and right now, we're pretty tapped out with what we've got in space. If we continue to get more comfortable with space engineering and we continue to get better at putting stuff up there, I wouldn't be surprised we start seeing some spin-for-gravity relatively soon now.

This is important because even things like convection in the air stop working well at 0g, so for example, you can suffocate on your own CO2 if you fall asleep and there isn’t a fan running.

Funny enough the hardest part about this doesn’t even seem to be the tethering or rotation, but two things; 1) you have to make slight course corrections along the way because the initial burn is never precisely accurate enough to target something as far away as a planet, and 2) occasionally you have to put as much mass as possible between you and the sun due to solar storms spewing radiation, so you have to stop the rotation occasionally to point the engines sunward to use your fuel as a radiation shield, or to course correct.

The challenge in realizing it almost entirely an issue of getting materials into orbit, which is the problem space tethers are targeting. The cost of getting to geosynchronous orbit is around $30k per kilo, so at the moment economics is limiting us rather than tech when it comes to building crazy space stations. If we could bring the cost per kilo down 100x or so then so many more possibilities open up.

There is a size limit, though--the required material strength scales linearly with the radius. Note that if you'll accept some fancy engineering this can be avoided: Instead of making it self-supporting you surround your station with a non-rotating ring. You can make it as thick as you need to provide the required strength. Wheels can't take the speeds involved, the connection will have to be magnetic levitation. Yes, this means maintenance but so long as you have an adequate safety margin you can power down and take apart any given piece for upkeep or even replacement. Just don't take apart too many at once.

The math pencils out, but the station has to have a fairly large radius away from the axis of rotation. For example a station rotating once a minute with a radius of 500m produces 0.6g. This means that you are looking at a 1km diameter if the station is symmetrical, or possibly less if some sort of counterweight is setup.

Not at all implausible with current technology. Two habitats connected with a tether would do it.

My sense is that the most valuable information to come from ISS is of astronaut adaptation and long-term space effects on health.

Specifically this video: https://youtu.be/86JAU3w9mB8

tl;dr: it's trivial from a materials point of view, but hard from a "but you have to build it in space" point of view. We're still not great at building large structures in space.