I don't think asking a robotic arms control activist about orbital mechanics and aerodynamics is going to result in anything insightful.

Plane changes are very expensive if you don't have the aerodynamics to get cross range. So if you have a traditional triangular capsule, yes, you do in fact need a honkin' big engine (and more importantly, a big fuel tank...) to expand your "landing zone". But if you have wings, even if the L/D ratio is not exactly a high performance glider, you still have an immensely larger "landing zone" so you don't need a honkin' big engine. Thats why they invest mass and volume in those wings instead of engines, tanks, and a parachute.

Another way to phrase it, is without wings you pretty much land somewhere along the ground track you would have passed over plus or minus like ten miles. But with wings, you can get hundreds, maybe thousands of miles of cross range glide. Your orbit can be over Chicago but you can decide to land in Florida.

We've had the strategic nuclear triad for decades, this thing's certainly big enough to make a strategic quadrilateral or maybe one leg of the triad is getting too expensive and needs retirement.

If you have patience, perhaps for a doomsday weapon or strategic deterrent, you don't need wings, giant engines, or much of a parachute. Its obviously a sensor platform not a weapon.

So what kind of sensor platform exists where its cheaper to launch 1000 one terabyte hard drives in a NAS than to downlink all that stuff. Well, "strategic" photographic imaging, maybe some kind of wide bandwidth SDR... The kind of thing were you press the record button and a year later gather demographic strategic intel, rather than a handful of snapshots of one semi-tactical target. Like how many total radar units does Iran have, well, you listen to the entire surveillance radar band over the country for an entire year, then do a lot of analysis once it lands. Not for "gimmie a pic of that one target today" type questions.