Woohoo! Only slightly sleep-deprived? I'm not going to post my sleep schedule, I will be chastised. And I'm a big boy, so I do my own scolding these days.
We're possibly entering another grand solar minimum. Very little storming/spots, the previous sunspot cycle was a dud, but it's almost over. No one really knows what happens next. Either way, we have been steadily putting more and more satellites up there, and another major coronal mass ejection event could happen any day, even when the sun's looking chill. We know space weather, but not space climate. If we have another Carrington Event sometime soon, I'm worried somebody will think somebody else has plans to use the opportunity for a nuclear strike, and go on the defensive (read: offensive). Regardless of whatever chain of events it sets in motion, we have the potential to see tens to hundreds of billion$ in damaged satellites. Yessir.
Geo is in the outer rad belt, true, but the outer belt is electrons, and the stopping power of several MeV electrons isn't so bad. You could (and probably would) have people living on the thing, underground, but not many, and they should have an evacuation plan that takes less than a day to execute, in the event of a "serious enough" storm. You could kinda control the asteroid's potential, if you needed to, with the ion thrusters, and you could shield the most sensitive electronics using the asteroid's surface geometry (and going sub-surface). As for induced currents from geomagnetic storming, I think you could design for it, but there's no need to run the thing during the rare storm. Strong currents from storming most intensely affect systems above a certain scale size (as I'm sure you know). Cosmic ray degredation of circuitry edit: and people :( might be the biggest problem, no atmosphere. No cosmic showers! Probably not many showers on that hunk of rock at all.
Here's another space-y idea for you to check out: Ballpark estimate how energy would have to be dumped into the Earth's atmosphere (over some span of time) to de-orbit the Iridium satellite constellation (via enhanced drag from more collisions cuz higher density) in a period of 2 hours? In a period of 8 hours? You can assume homogeneous heating of the atmosphere and equilibrium with a magical heating element that we're distributing evenly across all of the atmospheric particles, so still a maxwellian distribution (in the thermosphere). Edit2: I'm thinking it's not exactly maxwellian up there, but whatever, justify some boost in the distribution function/shape. You'll need a couple/few other simplifying assumptions, probably. I think in reality, heating would be quite localized to the auroral regions, and the Iridium satellites are in polar orbit, so they do pass through there. Actually, people get magnetometer data from those guys, AMPERE, and nobody tells anybody else about much of this cool stuff, because I dunno.
This is now a space thread.