Well, I’d say the real reason it’s not done is because (many) satellites are much more useful the closer they are to Earth. Communications links have higher effective power and lower latency, and optical/radar systems get better resolution.
The current sweet spot is about 400-500 km, which has enough drag that a satellite needs reboosting or it will deorbit in a few years. But low-thrust ion engines are sufficient here. Soon, we’ll see some constellations of very low orbit satellites in the <250 km range. They need more thrust and at that point need to be aerodynamic. But by several metrics they’ll be 4x better at their job than a satellite at 500 km (due to 1/r^2 scaling for radio, etc.).
I’m not sure how much this buys you for a nuke in space, though. It’ll shave something off the payload delivery time, but it might be negligible.
The downside of distance is warning. A geostationary satellite, for example, is 22,300 miles away. Even at 17,000 miles an hour directly downward (a heck of a lot of delta-v 90° to orbital velocity) the target can see it coming for over an hour; that’s plenty of time to launch the first, second, and third wave of ICBM’s.
Would an orbital platform have any advantage over ICBM’s in terms of warning? The delta-v required to simply “drop” straight down is huge. Reentry for most spacecraft tends to fire half an orbit away and rely on air brakes; true, this is to minimize the energy needed, but a high-powered de-launch from orbit would be easily noticeable too. In the long run, I don’t see any advantage to orbital platform over ICBMs except to make them an easier, less hardened albeit harder to reach target.
Orbital nukes only make sense with either a sucidal hair-trigger lukewarm “cold war”, or a single superpower so dominant they can prevent other countries from having a strike-back policy. But then, you have the same problem as Vietnam - what do you do to enforce your rule? Randomly dropping a nuke on a city seems overkill and non-productive, and you still need an intermediate forces to deal with guerillas etc.
Whereas a dueling pair of nations with orbital nukes would be a real nightmare - which both the USA and USSR realized.
AFAIK, the real winner in attack warning is the depressed-trajectory submarine launched missile with a point-blank range. Warning time in the single-digit minutes. You can’t even scramble your bombers or complete the entire command sequence to authorize and launch ICBMs in that time.
Depressed Trajectory SLBMs: A Technical Evaluation and Arms Control Possibilities Science & Global Security, 1992, Volume3, pp.101-15
Orbital nuke platforms are the opposite. Useless for counter-force. Pretty much only good for attacking things that can’t move out of the way, like a city
Nuking 'em from orbit is really only a viable strategy when you already have something in orbit with the capacity to nuke 'em. As has already been established, it was a strategy examined and rejected during the cold war because there isn’t much point in building an orbital platform compared to suborbital nuke delivery mechanisms.
However, if an orbital platform already exists, say because it’s the starship you arrived on, this strategy makes a lot more sense. It was proposed in Aliens because their primary objective (secure the colony site) had already failed. This strategy was also used to great effect in Independence Day if you change “orbital nuke” to “low altitude directed energy beam.” A more realistic alien invasion by attackers with the same goals probably would have had them firing energy beams or dropping bombs from orbit. They seemed interested in bringing human resistance efforts to a quick end, so wiping out population centers with overwhelming force is a reasonable tactic.
Sure, at geosynchronous a launch would take way too long, but you don’t need to get anywhere near geosynchronous height before orbital decay is negligible.
There’s been some work in air-breathing satellites. We’re still talking a near-vacuum here, but the satellite takes the small amount of available air (that would normally be purely a drag force), and accelerates it electrically. The satellite needs no net propellant–just solar panels. So it could stay at a fairly low altitude (a couple hundred kilometers) indefinitely.