The majority of the linked document details the purpose of the system. As far as worthwhile it's pretty clear that Rhixomatica and the villages in Oaxaca think it's useful:

"The solution we are implementing is inspired by the actual social and physical structure that we aim to service: each village in Oaxaca has their own fully independent core network stack, and each community is fully in charge of their own infrastructure. There is no central authority governing across communities, by deliberate choice. Because the infrastructure is operated in remote rural areas, often from a pole on a hill crest running on solar panels, and with directional wifi over large distances, network links between villages can be unstable."

A valid response is that as soon as a data service uplink is provided, then common IP routing via the GGSN could take over, and everyone could be using messenger apps. D-GSM (btw not limited to 2G) is indeed most interesting for routing "old style" voice calls and SMS. These are still in wide use, and arguably more stable than pure data links, which have to compete for bandwidth with other data (base stations can actively prioritise voice).

If you think LTE (4G), then everything is already a data link at first sight; but most carriers still use CSFB, where the voice calls fall back to a 2G RAN.

The concept is important, as it structures the call flows in a new way, regardless of content (SMS, data, voice). The bearer (2G/3G, 4G, 5G) is much less important.

Once new architecture exists and is proven to converge at scale, then homing a subscriber data connection through one mechanism or another (GTP, PMIP, ..., wireguard?) is just a technical choice.