Each mobile generation was sold on speed. That framing is losing its usefulness, because current networks are already faster than most applications need, and the next generation is being designed around different goals.
What 5G delivered
Faster connections, mostly noticed in dense areas where earlier networks congested. Lower latency in principle, though few consumer applications used it. Considerable capacity for fixed wireless, which replaced cable in some markets.
What it did not deliver was the transformation promised in the marketing. Remote surgery over mobile networks did not become a thing.
What the next generation targets
Sensing. Networks that use their radio signals to detect position and movement, turning the infrastructure into a sensor as well as a channel.
Integration with satellites, so coverage does not stop at the edge of the terrestrial network.
Energy efficiency, because the power consumption of dense networks has become a genuine constraint.
And reliability guarantees, for applications where a dropped packet has physical consequences.
Why timelines slip
Standardisation is a multi-year negotiation between manufacturers, operators and governments with different interests. Spectrum allocation is political. And the business case for operators, who spent heavily on 5G and saw modest returns, is not obviously compelling.
What to expect
Deployment starting around the end of the decade, in dense urban areas first, sold on capabilities most users will not notice. The pattern of every generation so far.
The interesting applications, if they arrive, will come from the sensing and reliability features rather than from bandwidth – and they will take years to appear after the network does.