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 lack of open public spectrum is a clear mistake of the governing authorities. To run a GSM+3G+LTE network for the annual CCC congress, we have to ask commercial operators for ARFCNs to use, every year all over again, and every year it is unclear what the response will be, often up to very near the build up of the event. If we had a publicly usable spectrum for 2G,3G,4G (and 5G?) here in Germany, so much more innovation and free/open infrastructure projects would happen in mobile communication. Network coverage holes are high up on the news these days, just imagine if locals were allowed to help themselves and not rely on profit-only oriented commercial operators...
Germany's new "campus network" spectrum regime is the first step towards this. It isn't true free access for everyone, but it makes approx 90 MHz of usable mobile spectrum in the ~3.5 GHz band available.
The Netherlands has had around 5 MHz of DECT guardband spectrum available for private use, which is also useful. The UK has announced a new way to "share" access to 3.3 MHz of DECT guardband (plus 10 MHz of 2.3 GHz, and any unused mobile operator spectrum).
Coverage blackspots are definitely a problem, but with these moves, maybe we can see some initiatives like these helping? The biggest challenge is likely in getting incumbent operators (paying billions for spectrum) to work with people willing to build these kinds of networks - until that happens adoption may struggle, and it is hard to onboard users and create good business models. Spectrum was definitely a barrier but in rural areas in northern Europe, hopefully this will be less of an issue going forward.
re "strange to choose 3g": as some operators even phase out 3G completely, we can expect 3G base stations to be available cheaper than ever. That's only one of the reasons: Osmocom started out from the security analysis perspective; having 3G openly available is a benefit to the general public as well as small business in various ways.
(The opening lines of https://osmocom.org/news/30: "3G is dead, you may think. From the perspective of large scale operators, that may well be the case, but this is precisely the reason why Open Source support for 3G is becoming increasingly interesting: when the focus for earning money shifts towards LTE infrastructure, the threshold for setting up 3G networks is becoming easier to surpass for everyone else.")
3g operates on licensed frequencies (sold for billions in some countries[1]), so you won't be able to setup a functioning network anyway. Maybe in some 3rd world country, in the middle of nowhere or a tiny femtocell..
And this is the thing that cheeses me off the most.
Look at the economic activity created around a "junk" wireless band: 2.4GHz. It is junk because of absorption by water molecules, and that's why microwave ovens use it.
Despite how crappy it is, and how much interference it gets, it is so, so useful. WiFi, Bluetooth, and dozens of other protocols. We get so much use out of it.
If some other bands were opened up in the same way for unlicensed use (conforming to a TX power limit of course), we could create more and more applications, especially if it was a bit lower in frequency to allow longer range communication with reasonable power consumption.
But there's no space worldwide for that, because it has all been sold off or otherwise provisioned. It is such a shame.
True and the German government was very "wise" to auction every last arfcn of their spectrum so no one will attempt to do teaching or R&D in Germany.
But not every country tried that hard, e.g. in the Netherlands one can broadcast with low power (e.g. have your own indoor/campus wide network) or in the US spectrum can be licensed for smaller regions.
Functioning can mean a lot of things. A security researcher trying to downgrade a device from 4G to 3G and then to 2G will be quite happy with a single eNodeB/nodeB/BTS set-up and is glad about low output power. A company to provide coverage on their campus can be quite happy too.
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.