The cost of deploying a significant wired infrastructure was, and still is, enormous, if you want to cover a large area and serve many people. Actually, if you are not deploying an infrastructure that reaches many people and you don’t have many people using it you won't be able to attract more people. It is the classical "the chicken or the egg" problem. If there are not many people using the infrastructure you don’t have any incentive in deploying it, on the other hand if the infrastructure is not pervasive and serving many people I, as a prospective user, don’t have any incentive in connecting to it. This has been proven over and over by ITU, the International Telecommunications Unions, in its yearly report on the status of worldwide telecommunications. It took a Country roughly 19 years to move from a 1% to 10% penetration, 10 years to go from 10% to 20%, 7 years to go up to 30% and 5 years more to reach 40%. As more people are using the telecom infrastructure more people get interested in using it. This is summarizes as the Metcalf law: the value of a network is proportional to the square of the points it connects (people using it).
Moore’s Law has fueled the deployment of infrastructures only marginally, because most of the cost was in the laying of cables and in the civil engineering related to that. And in this area the Moore’s law has no saying. It helped a bit in decreasing the cost of the service (switching) and the operation cost (digitalisation made possible to pack more calls onto a single wire, thus diminishing the need for expanding the infrastructure as more capacity was needed. However, it made terminals, like fax machines, answering machines, cordless phones more affordable and thus expanded the demand for the infrastructure and led ot increased revenues.
The digitalisation of the last mile, however, was basically the result of the Moore’s law. The increased processing power has made ADSL possible and the decreasing cost of chips made it affordable. The Web would not exist without the capacity increase and cost decrease sustained by the Moore’s law and the iNternet would have remained a connectivity for few universities.
The ADSL is important for Smart Cities as a technology. The possibility to code a digital signal into a set of frequencies can be twisted to operate on other kind of infrastructures, like the illumination grid of a city or like the city power grid. It is not exaclty ADSL but the principle is the same.
This means that a Municipality may consider leveraging on its own “utilities” to deliver and harvest bits at a marginal cost.
Speaking of cost, the laying of cables, both copper and fibre, remains very expensive (even if the fibre is cheaper than the copper, the civil works needed for its deployment are high and tend to grow over time). However, the wired infrastructure remains (and will remain for the foreseable future) the best way to carry a lot of “bits”. The capacity over a fibre is at least a thousands time bigger than the one provided by radio and it is also at least a hundred times cheaper in terms of power.