Indeed it is not a telescope mirror, and yes that makes matters much easier. But still far from easy. The tech to maintain orbital position and attitude for telescopes and comm sats is well established; now multiply that by... hmm, let's see.
If we want to double the incident light on Mars using a whole bunch of mirrors 5 to 20 degrees ahead of or behind the planet then we need a total mirror area about the same as the planet's presented area (by which I mean the area of its apparent disk; I don't know it there's an official term for this.) Mars's diameter is about 4200 miles, so it's presented area is almost 14 billion square miles, or
155 billion separate reflectors if each is is a 50 foot square.
Station keeping and attitude control for one of these would not be very difficult. The electrical power, physical structure, and fuel requirements for 155 billion of them would undoubtedly be far greater than those of many structures each comprising many of them, let's say (pulling a number out of my backside) 155 million assemblies of 1000 reflectors each. Roll that over in your mind for a moment: each unit has nearly a tenth of a square mile of mylar and there are 155 million of them!
The design life of a typical GEO com sat is 15 years, and the fuel load is enough for somewhat more than that just in case the equipment hangs in there. Many do last longer; many, not all. But notice, it's not the fuel load that limits life, but rather things like equipment failures in the harsh thermal and radiation environment and degradation of the solar arrays from those same factors and micrometeor damage, among other things. So, let's just suppose our 1000 mirror satellites can do better on life; let's suppose they do lots better and manage 50 years. To replace 155 million satellites in a span of 50 years means manufacturing and positioning 3.1 million of them per year, or about 8500 per day, 5.9 per minute.
Can we build and launch 5.9 of these units and the launch vehicles to lift them at a rate of 5.9 per minute? Maybe we could do multi-unit launches, so we only need one launch vehicle per minute; that's a whole lot better
. And this isn't the initial deployment, this is the
replacement rate for the full-up system, which means it never finishes.
I'm not even going to try to estimate the total mass of various materials - aluminum, silicon, copper, various polymers, fuel, etc. - being consumed by this and sent off planet every day.
And all this just to double the incident light on Mars. Which would bring its total incident light
almost up to that of Earth.