I don't see why we could not just build a factory on the moon for making foil. The first hurdle would be making Fusion a reality. The moon's crust (and dust that you want to 'vacuum' away) are ruch in Helium 3, which is essential for fusion. Since it is just 'lying around' it would be relitively easy to move it to power generation. A bonus, is that one person walking around with a shovel would probably be all the manpower you need, as a metric ton of helium 3 would be enough to power the US for a year.
This huge amount of power would be required for the sepearation of aluminum from boxite, which requires HUGE amounts of power. In fact, most aluminum plants are built right next door to a power plant, and all the power plant does is supply the aluminum plant with power.
Reynolds plants make between 200 and 350 thousand tons of aluminum per year which is about 120 million cubic meters, which equates into about 10 plants for a cubic kilometer if we wanted to accomplish this feat within a year. Supprisingly, that is slightly less than how much aluminum Canada manufactured last year.
In order to get that amount of foil applied to the surface of the moon in a decenly efficient process, the foil would have to be modified. The foil would have to have reenforced edges in order to withstand the stress of being unrolled quickly. This could be done by modifying the rolling press to quadruple the tickness of the edges. The foil would basicly have to look concave when viewed on edge as any quick changes in thickness would lead to stress points which would tear the foil.
Now, current aluminum plants make foil in rolls that are less than 3 meters wide. If we scale up the size of the rolls to 10 meters wide we would have less of a problem with waste, since now we do not have to overlap the foil as much at the edges to make sure that the surface doesn't show though in rings. Having a striped moon would suck. In additon, a 10 meter wide roll is just about at the limit for easy alignment from an operators standpoint. (reference - farming)
Assuming we set a goal of 10 years for this project, we would need to unspool the foil at roughly 12,000 meters per second; assuming start from completion of the factories and fusion plants. The capacity of 2 plants would be more than enough to keep up with the unspooling rate, but the unspooling is the bottleneck, so we ould want enough capacity at the plants so that if a plant goes down, the unspooling process could continue with increased output at the other plants.
In order to get 12,000 square meters out per second, a 10 meter roll would have to be unfurled at about 1,200 m/s or about 2200 miles per hour. This is unacceptable from a ground point. While it could be done using a satalite unfurling foil, the logistics of changing a roll in midflight nearly daily would be a nightmare, along with having to change the unfurling speed in milisecond intervals to account for ground height variations. craters can be steep, and a motor on the satilite would be unlikely to keep pace with the contours under it.
A better plan would be to use multiple ground vehicles to unfirl the foil over flatter terain, and use small crews to do steeper secions. A 3 man crew and vehicle would be able to transport a roll to a crater, and basicly roll the foil down a steep crater side to the bottom and secure it. This would be the most labor intensive piece of the project.
Probably the best way to look at the logisticls of covering the moon is to assume it is flat, and take the crews out of the math when figuring how many vehicle teams we would need to cover the moon. Now, this is not entirely true as the logistics of getting a ridge covered before a vehicle team gets there to do the flat areas would have to be worked out, but that can be done by adding more 3 man teams to the steeper secions.
Assuming that, and assuming a safe and sane speed of 36 kph on the moon (22 mph; remember we are driving on what amounts to dust with electricly powered vehicles) we would need 100 vehicle teams unfurling 24/7/365 for 10 years to get the moon covered. Obviously, we would need more than 100 teams, as even slave labor from Cananda would have to sleep. If we are good masters, we would require 10 days work time, 2 15 minute breaks and a 1 hour lunch. This takes our workforce from 100 to 240, but I believe the downtime is worth it since we can not expect many children to be born and usable durring a 10 year span to replace overworked workers.
Therefore, we wold be looking at about 15 years to cover the moon in foil. 2 years to scout and plan, 2 years to build the factories and infrastructure, a year of built up raw materials, and 10 years to lay the foil.