Suppose that you can build a reflector surface with a 10 micron aluminum foil (27 grams per square meter) and an additional 13 grams of support mass to keep it

Suppose that you can build a reflector surface with a 10 micron aluminum foil (27 grams per square meter) and an additional 13 grams of support mass to keep it rigid/oriented, for a total mass of 40 grams per square meter. Near Earth but outside atmosphere the solar constant is about 1366 watts per square meter. That puts an upper bound on how much warming you can prevent per m^2; in reality the Earth's atmosphere and surface already reflect a fair part of that energy without absorbing it. If you send the reflectors to geostationary orbit they'll be in sunlight constantly, but I don't know if their full penumbra will actually fall on Earth, plus that reduces the deliverable payload mass. Instead let's make another unrealistic simplifying assumption for the sake of argument: pretend that the shade array goes to LEO but it's reflecting sunlight away from Earth 100% of the time. At 40 grams per square meter of reflector and 1366 watts per square meter that's an upper bound of 34.15 watts per gram or 34.15 megawatts per tonne of avoided insolation. With one expendable launch of 500 tonnes to LEO that's an upper bound of 17,075 megawatts of avoided insolation per launch. In 2007 the IPCC estimated the net anthropogenic component of radiative forcing for the Earth system at ~1.5 watts/m^2, or 765,108,000 megawatts altogether (1.5 * 510072000 * 10^6 watts). Using the assumptions previously given, one maximum-payload launch could mitigate about 0.002% of anthropogenic radiative forcing. Adjust the achievable mitigation upward proportionately if you think that the shield system can be implemented with less mass and adjust it downward if you want to replace my optimistic upper bounds assumptions with more realistic (lower) ones.