If the habitat is segmented and connected with non rotating couplers then you could mount solar on each coupler. The design would then be multiple identical potentially self sufficient modules that can be assembled into a large colony and expand as needed. Wouldn’t have much of a horizon though so it would feel a bit claustrophobic.

Why bother connecting the solar arrays to the habitat? Just hook them up to microwave transmitters and beam the power to the cylinders. That way you can have the solar arrays anywhere it's convenient, maybe even in position to shade the entire colony so that the heat disposal problem is less of an issue.

Fair point but you needn’t bother with microwaves when a cable and some good ol’ fashion brushes will do. It’s just a rotor after all.

Maybe it would make sense to set up the solar array like a huge umbrella on the sun-facing side of the colony. Then you can cover the now-shady outside surface with radiators.

I assume we break up the colony into at least two segments so they can be counter-rotating.

Forgive me, what is the benefit of having counter-rotating segments? Does it reduce the overall angular momentum of the object...?

No, but it prevents precession IMO.