There are possible orbits that get full sun (L1 orbit for instance), but the earth still rotates underneath, so beaming the power down (typical studies use microwave beams targeting large (like 1/2 of Nevada size) antenna fields) is problematic. People might not like the extra warmth of those microwaves beaming down on their heads in an already hot area (maybe only a problem at the edge of the antenna grid?), or they might be unhappy with a large wire grid network spanning thousands of square miles overhead.. Cost effectiveness has never really been studied, the cost to lift enough panels to make it worthwhile has been too high historically. It might become cost effective if a lunar base ever gets established. Then you need to consider the maintenance, solar cells degrade in space due to mechanisms not found on earth - i.e. solar radiation ion damage and micrometeorite ablation.
From memory, it was awhile ago I worked with colleagues studying the problem for comsats, but it is theoretically possible to get about 90% of 24/7 power in a geosynchronous orbit, since Earth's full shadow zone is only a few hours of the orbit, and that only in the parts of the year when the axial tilt of the satellite orbit is aligned with the earth's orbit around the sun. So, 24/7 with battery storage on orbit...or multiple satellites so you can have 2 out of 3 beaming at "night".
"The poles definitely wouldn't work". Um, no Josh. There are a couple ways to make a near-stationary polar orbiter, one that might work for a power satellite is one of R, L. Forward's statites. Again, not been demonstrated, and certainly more expensive than a geosynchronous satellite, but not impossible. Orbital dynamics is a fun subject.
What if it was located over one of the poles?
