I'm with Jannie on this one - the article is misleading in the extreme. While a cool project, it is not quite a new idea, or a replacement for a BTS.
1) The height of the tower has nothing to do with the BTS size. If you want to replace a tower, show me that their system provides coverage at the same cell edge as that of the tower.
2) The BTS is but a small part of the equipment in an MCO cabinet - most will contain one or two routers and backhaul equipment.
3) OpenBTS as a project has been around for a long time now - the only thing that they did that is new is to implement it on a Pi instead of on an X86 computer. The OpenBTS implementation on X86 hasn't really succeeded in making much of a dent on the market, so why should the Pi?
4) OpenBTS supports only GSM. Most modern BTSs supports GSM, 3G and LTE, or is upgradeable to LTE.
5) This implementation of theirs basically plays into the space of a
femtocell, not a fully fledged BTS.