I think there are a lot of plausible ways this can happen.
The core idea I think is, it's like colo/dc vs aws. Both make sense in different circumstances.
Higher pricing for flexibility, elasticity - avoiding lock-in while scaling up or down (maybe to 0) or moving location. Price transparency.
Network effects, brand recognition, standardised services and high quality "overlays" to spaces like wifi, printers and presentation equipment that works everywhere.
Social networks - in physical space, and improved discoverability of spaces themselves and related business clusters.
There is also a scale play where you get big enough as an intermediary to beat up suppliers - because you can control and direct demand. Usually a landlord gets to extract surplus value from a location but as WeWork's overlays / brand / network grow powerful they have a stronger negotiating position. Consider the situation where WeWork is huge, and having one near other properties you own (housing, food, retail, etc) raises the value of those.
You could take this further and who knows if WeWork will be the ones to do it, but there are definitely value improvements to be had over "negotiate with an agent and commit to an n-year lease, then do a fit-out". Serviced offices are a thing but scaling that idea _could_ be something else again.