The Amazon example is less relevant on that specific point as AWS owns and operates its own equipment.
I don't see why it would be any harder than selling off a plane in case of market downturn. Both are hard, and in both cases the owners of the asset (whether it's a lease or a plane) will likely incur a loss, so I don't really see any difference here at all.
You are really pushing this point hard, when it does not fit.
No such lock-in exists with WeWork. In fact the entire reason for WeWork's existence is it doesn't have lock-in.
This isn't true for their expanding customer base: Businesses who are paying for blocks of co-working memberships to expand through services rather than finding new leases themselves. There is significant lock-in when this occurs, as people may move close to those WeWork locations, arrange their life around it, etc.
WeWork might be able to survive long term on two factors
1. Prestige & Environs - working in this sort of shared space may be viewed as luxurious in the long term due to rubbing elbows with other well-to-do folks, sorta like going to the opera.
2. Reliability of cost - Owning a condo is a strictly better decision than renting an apartment, but there are spikes in costs when appliances break down and building work needs to be done... though I think to properly reap the benefit of this WeWork needs to be and remain extremely proactive in maintenance, if some vulture capitalists get into the decision making process and let maintenance slip for a bit it'll erase all of this value by forcing the cost (in terms of inconvenience and unavailability) onto the consumer - it's possible this could also be lost if they fail to keep a nice margin of capacity above usage... again, as soon as someone wants to go to work and is unable to secure the space and is forced to either rent or lease conventional office space their competitive edge will dissolve.
I am highly skeptical of the long term prospects if WeWork fails to maintain their prestige and banking your company on prestige is an inherently risky approach IMO.
Every airline that takes a loan to buy a plane is banking on future demand for air travel, for example.
Or take Amazon Web Services, which is building data centers to lease out on a short term basis.
Yes, there are tremendous risks involved, depending on the lease terms. But it's not at all unusual.