[petra]

It seems that something like Loon is much more risky and large scale than those Bell Labs projects - which seem like regular r&d risks (although many of them).

And if it have worked reliably and affordably, they probably could have found some way to monetize it.