Your math is close for a total earth disk sized shadow. The proposal is to block 1 or 2 percent of the total insolation that would be a disk 150 miles in diameter for a 1% interception at L1 with carbon nanotubes and spinning the disk it could be literally atoms thick to block the IR light wavelengths. Solar photon thrust would be a very real issue you have just created a solar sail on a scale that has been proposed to send probes to proxima Cent. The closer star to terra. You would need to tack the sail against the solar wind to put the thrust vector against the wind vector otherwise it will.rapidly accelerate to sol C3 escape velocity. The theory is valid and a sail of a few hundred miles in diameter is within human technical reach. Should it be done is the better question at least if they messed up the calculations and it cooled too much just tilt it with the solar wind vector and away it fly out of the solar system forever.
A solar sail of somewhat higher density but much smaller size would cast a shadow that would block only a small part of the solar disk (viewed from Earth surface) but would have to block a lot more of the solar radiation per unit area, and thus would be subject to much higher degradation. The Solar Wind is not to be trifled with. And then as you point out, you have to keep it from blowing away.
The shadow would necessarily oscillate +/-22 deg in latitude with the seasons, and probably piss off a lot of solar panel enthusiasts. Such is life.
BTW, in the 80's I designed attitude control systems for LEO and GEO spacecraft, using combinations of IR limb sensors, gyros, and electromagnetics, but those birds were orbital and had entirely different missions. Thinking about the problems of placing a quasi-stable sun-shield at L1 fascinates me.