This take makes a very clear argument, but I think it's important to question that, as a company grows, it is generally "willing and able to deal with building owners directly" - a similar argument could have been made for the early days of AWS, when it seemed like a great way for startups to avoid managing low-traffic, scalable infrastructure, but redundant for companies that had the resources to manage their own servers. AWS has shown how successful a 'platform middleman' can be, even for very large companies.
WeWork is betting on taking a similar track - sure, companies _could_ decide they want to build out expertise in leasing and managing real estate space, but it can actually help them focus on their core business more by leaving it to another company that provides the 'platform'. The growth of WeWork's enterprise offerings and general awareness of how to pitch to mid- and large-size tenants indicates that growth projections are not pinned to a startup-driven economy.
For me the big difference between AWS and WeWork is that a company will always have people dedicated to figuring out their offices. WeWork isn't solving problems like "who do we have, how do we arrange them, and what kind of spaces are best for how they work". On top of that, finding and negotiating leases every 5-10 years isn't a lot of extra work. AWS in contrast eliminates entire classes of work.
Although now I think about it, perhaps demand volatility is even more important. Office space needs are way more stable and predictable than server load. Salesforce, for example, signed a 15-year lease for Salesforce tower. And there's every reason to think they'll use it. Whereas AWS and its customers benefit hugely from being able to change load both month to month and hour to hour.
WeWork is betting on taking a similar track - sure, companies _could_ decide they want to build out expertise in leasing and managing real estate space, but it can actually help them focus on their core business more by leaving it to another company that provides the 'platform'. The growth of WeWork's enterprise offerings and general awareness of how to pitch to mid- and large-size tenants indicates that growth projections are not pinned to a startup-driven economy.