I doubt it's never going to be practical enough to be useful, at all. I might be wrong and someone can do the calculation to prove me wrong, but I don't think

I doubt it's never going to be practical enough to be useful, at all. I might be wrong and someone can do the calculation to prove me wrong, but I don't think there is any way you're going to power an army or even a single regiment with that (or even more than a couple of toasters). First, Low Earth Orbit is at least 160km above the Earth. You could do it at 100km but now you have to fight atmospheric drag on top of everything else and constantly need to readjust your orbit to avoid burning up. Forming a beam tight enough that it stays focussed on your small target is the first challenge. Talk about fraction of a degree. You'll get much worse energy loss just there, because your target will always be smaller than the energy spot. If you have to build large targets, then there is a point where it would be more advantageous to just use solar panels on the ground. Second, since you'd be using a low orbit, you'd still be in the dark at night, so no power for you. Third, the lower the orbit, the faster the satellite is going to zoom past in the sky. You'd need a constellation of these to constantly reshape their beam to all their targets. As their distance from the target increases, the power received will also vary a lot. It also means the satellite will have to know the precise location of the target at all times. Might be workable for a static unit, but if you're talking drone/autonomous system, then they will need a way to transmit this information somewhere, very, very accurately too so the beam can find it. It's not something you could deploy anywhere in an instant unless you have thousands of satellites covering all areas of interest. The solar panel size you would need to ensure some decent minimum power received on Earth would be pretty big. Say you have a super efficient solar panel that can get you 30% efficiency, you'd get max of 300W/m2. So you would need very large arrays of panels just to get a few kW on the ground. Now if instead of solar panels you want to beam up energy to bounce from the satellite back to the target, you'd double your problem space: need lots of large emitters that would need to target a constellation of fast moving satellites. Not great unless you know you can deploy them exactly where they are needed, which probably limits the use cases in foreign territories. So, I'd say this will remain a niche application on Earth, where the size of the transmitting arrays depends on the distance to the target, and will probably never go beyond a trial in space if they manage to convince someone to get funding for it.