[1] Caveat Starship has to reach its goal of transporting 100 tonnes to LEO
A bit off-topic, but an aerospike engine is half of a rocket nozzle, with a virtual half created by the supersonic shockwave. So we could envision a retractable nozzle half that moves through subsonic, transonic and supersonic modes to power the airship.
Also the SABRE engine uses (according to AI) 16,800 thin-walled tubes filled with liquid hydrogen to cool ambient air to -238 F (-150 C or 123 K) in 10 milliseconds so that it can be compressed up to 140 atmospheres and fed into a combined-cycle engine. That would allow it to be air-breathing up to mach 5.4 (3,600 mph or 1.6 km/s) and transition to liquid oxygen after leaving the atmosphere.
I also asked it about using something like titanium to withstand the heat of exiting the atmosphere (since the titanium SR-71 reached mach 3+) but it said that it can't withstand a high enough temperature. So an ablative coating might need to be applied between launches. Quite a bit of research was done for that through about the 1970s before NASA chose the space shuttle with its reusable tiles.
It seems like most of the hard work has already been done to achieve this. So I don't really understand why so many billions of dollars get devoted to other high-risk ventures like SpaceX. When for a comparatively smaller amount of money, a prototype spaceplane could be built. I'm guessing that the risk/reward value just wasn't proven yet. But really shouldn't VC money chase the biggest bet?
This is the kind of stuff that I went down rabbit holes for when I dreamed of winning the internet lottery. Now that AI is here, I can feel the opportunity for that slipping away. A more likely future is the democratization of problem solving, where everyone knows everything, but has little or no money and doesn't want to pay for anything. So really not much different from today. So maybe it's better to let these half-baked ideas go so that someone else can manifest them.
Needless to say, getting anything to go to space is hard.
A nozzle engine doesnt have to account for this as much because the nozzle is keeping the pressure of the exhaust
If the pressure at exhaust is higher than ambient, the exhaust pushes outward against the ambient pressure and you get huge exhaust plumes, and lost efficiency.
Conversely, if the pressure at exhaust is lower, the ambient pressure pushes the exhaust inward into shock diamonds[1] and you, again, lose efficiency.
Engine bells specifically yield their max efficiency at one external pressure/altitude. The reason you see shock diamonds is most often from ground-level testing (or takeoff) of engines that perform best at altitude.
You should be for example totally able to build an inflatable frame for a launch loop, making it possible to launch payloads from above most of the atmosphere.
Also for re-entry the more you lower the density of something, the less re-entry stresses there will be. So you could construct a giant low-pressure inflatable decelerator device and have it essentially float down all the way from orbit, incrementally shedding energy as it comes down over a longer period of time, taking care to balance the rate of descent, heating and internal/external pressures.