If we use tethered balloons with wireless gateways fitted on them (let's say for Cellular or LPWAN communication), they can provide much more coverage. Why aren't they used then?
While there are many instances of high-altitude balloons (like Google's Project Loon, Space Data SkySite® or this question on this StackExchange), I could find scarce resource on low-altitude tethered balloons.
What I could find was this work done in 2008 (IEEE Conference Paper) by a research lab. Another recent conference paper (2016) is this from another lab. The first paper also demonstrates that the tethered balloons turn out to be lower cost in the long run.
I wonder how come we don't have tethered-balloon based wireless gateways when they are cheaper and provide more coverage than fixed towers?