"Alternate Food Sources' reminded me of the terrible movie with Rob Schneider hiding inside a service robot.... "Eat recycled food! It's good for the environment and okay for you!"
Thoughts:
1. How big of a system ship can you land in your TU? That impacts shuttle efficiency.
2. Do you use a RO-RO model / LASH sort of scheme - if you do, shuttle loading might be very efficient....
3. Some sci-fi uses magnetic acceleration to launch cargo pods into orbit or to assist orbital shuttles heading upwards (kick start)... not sure what that costs, but it could reduce costs to orbit which will be more expensive than costs from orbit (assuming you have an atmo and can use wings, flaps, etc).
4. If the cost of disposable or one use re-entry chutes was cheap enough, you could use steerable ones to drop cargo containers to the surface. If the containers were recyclable (turn into building materials or degradable), then that might be a cheap way to get cargo down.
5. Traveller posits flying cities somewhere around TL13-TL14. To do that, you need to have a) mostly killed off any sorts of resource scarcity (the materials involved in building a large hab that continually floats would be insane) and power must not cost much (or you would just not do this). So if I can do a floating city, a beanstalk seems feasible. Charles Sheffield and other authors have suggested they are doable without anti-grav, but if you have cheap AG that is reliable, that could help.
6. Traveller economics is whacked in some ways. Look at the cost of a nice aircar built at tech-15 - it can be very pricey. Ships built at TL-15 also expensive. And yet, in order to build super massive ships and floating cities and L5 habs, and whatnot, you'd need stupendous number of credits in a build system consistent with the ship and vehicle costs at TL-15. There is an argument that TL-15 places will be prosperous ... BUT.... freight still costs the same as it did anywhere if it comes in from offworld and speculative cargos can be worth more. And we don't see higher mustering out benefits for people returning to their higher tech homeworlds. To me, a lot of things, rather than getting more expensive as you go up in TL, should go down in cost.
The whole point a lot of the time is to make things of a slightly better quality for MUCH less cost. That's how a lot of progress runs in the modern world. Yes, inflation nibbles at that (a decent PC a 15 years or so back would have cost $2K and now could easily cost $3-4K with all the bells and whistles) and that also illustrates feature creep, but for the things that we've built that have been roughly the same, where resource scarcity was not a factor (something less and less true on our world but necessary for many big 3I projects), the costs have dropped. The only way to build megaprojects (space habs, floating cities, beanstalks) would be - good high tech materials at low cost (say very easy availability of stuff like carbon fibers or whatever futuristic possibilities for materials we don't know about yet) that are cheap, cheap power, and reliable and cheap large scale AG units, and powerful, cheap computer control systems.
So, if we posited a more sane setup to construction, a floating city should not cost more than maybe twice what a regular city would cost (otherwise maybe only the capital would ever get the ridiculous expense). And by comparison, a beanstalk might be less than a mid-sized city to build. You might even build to on opposite sides of the globe (or more if you spaced them out evenly.
7. In terms of operating efficiencies:
a) Downbound cars could use regenerative breaking along the say to generate power.
b) Outbound cars could in large part be filled with waste materials (for reprocessing in orbit or being shot into the sun) and their movement upward could be partially aided by other cargoes coming down (like a ballast on an elevator).
c) Having containerized shipping would allow a fast transfer of ship to beanstalk. That's an efficiency that could save you money.
d) If you believe you can build your beanstalks without AG (after construction where that would really help getting them in place), then you can save a lot in construction and you can save a lot in operating costs. This could make your beanstalk very cheap relatively speaking (esp if you imagine a more reasonable cost of higher tech projects).
8. Last point: Beanstalks are nice, but if they break and collapse... you've got a planetary scale disaster. In 2300AD, I think there was one on a colony world that was brought down by the Kafers and it wrapped around the equator doing a lot of damage. If you have beanstalks, I think I would not allow any building within X many kms of the equator to avoid massive death tolls if it eventually fell. Might make it 1000 km each way or at least 500 km each way. In 2300AD, I think they severed the counterbalance rock that helped tension the beanstalk and that spelled the end of the beanstalk. An AG stabilized version (even if the AG only kicked in during emergencies so wasn't an normal operating expense) could avoid a lot of that risk and give time for a repair.