Mobile signal strength, like Wi-Fi signal strength, have become bigger issues as frequencies in use have climbed from the hundreds of megahertz ranges up into the gigabit ranges.
That problem expresses itself both in terms of indoor and outdoor coverage issues, though the indoor issues arguably are the toughest problems to solve.
Some might, in retrospect, bemoan the fact that newer protocols such as Long Term Evolution were not built, from the inception, to gracefully support voice. That is a reasonable criticism, in retrospect.
The defense is that, at the time the standards were developed, the crucial and critical deficiencies of mobile networks were the ability to support higher speed Internet access. As always, there was not unlimited time to create a standard, and get it into commercial deployment.
Though that doubtless will change with 5G (voice and messaging will not be an afterthought), one surmises from most of the verbiage that voice and messaging, though parts of the standard, are not by any means the key problems to be solved.
Bandwidth, latency, application support and the core network features all seem to be much bigger issues than ensuring that voice and messaging “works.”
Small cells will help. So will bonding to fixed network bandwidth (Wi-Fi). But voice performance still will not be a huge issue for 5G. The function has to be supported, to be sure. But it might not get as much attention as you might previously have thought.