That’s just how research works most of the time. The experimental setup required to build a working prototype and prove the initial hypothesis is always going to be larger and more complex than a mass market appliance. If that appliance ever gets built depends on a huge number of factors too. If the process scales as expected, how complex the device is to produce and if a company thinks that it can make money on it. The researchers, meanwhile, are probably more worried about their next grant funding.
Solving the fragmented community problem is something I’ve been pondering too, and the meta-community idea you described seems interesting.
Obviously, a proper technical solution will be difficult. Federation comes with a host of challenges, as well as benefits.
Giving communities the opportunity to be open to other like minded people on different instances would be beneficial to the network, for a number of reasons.
If two communities on different instances have the same name, it doesn’t seem crazy to ask each of them if they’d like to federate with each other.
That way, apart from instances defederating, discussions could continue even if individual servers go down.
Of course, people love to hold on to their little fiefdoms, so the issue is as much social as technical.