Satellites already exist for data connections and are quasi profitable. However there are technical issues with satellites for broad band internet service, and the biggest is the available spectral space is quite limiting for vast tracts of the developed world. Basically the issues with satellites are more than just cost
But that begs the question here really.. Is this new approach of using temporary balloon based distribution with the effort? I'm not so sure. Where I see the advantage of this idea, how's the operating cost of such a system going to be less than the existing cell network? What spectrum space are they planning to use? How will they obtain the rights to that space unless they buy it? Perhaps they plan to use 802.11 A/N spectrum under what ever passes as FCC part 15 in the UK?
If they are buying spectrum, they will go broke before they begin because the will be bidding with the Cell companies for the same space... If they are using unlicensed 802.11 a/n, then I wish them luck but I don't support their effort. 802.11 a/n is congested bad enough now, we don't need more emitters floating around making matters worse. Plus as a Amateur radio licensee, I have rights to a lot of that spectrum and thus have priority over part 15 users, and I actually do use some of that space and would not welcome their unlicensed intrusion.