> Won't these other competitors have to purchase some bandwidth from the government as well?
There's tons of bandwidth in the millimeter wave spectrum where key 5G deployments will happen: https://www.verizon.com/about/our-company/5g/what-millimeter.... And the government is working to clear more.
Proponents of regulation have this tendency to view things through a moral lens. "Wireless should be regulated because it uses public spectrum." But the question isn't whether regulation is morally justified, but whether it is economically efficient. That means the focus is not on whether some public resource is being used, but how many competitors can be sustained with those public resources. Even with LTE, there are four nationwide competitors, which is more competition than exists in many other markets. And 5G can support even more. Whether or not those competitors need to use public spectrum to provide their service, there will be enough competition to obviate the need to regulate.
> The moat that Coke and Pepsi have is maintained purely by marketing.
What you call a "moat" is in economics terms a "barrier to entry." And there's lots of different ones. E.g. anyone seeking to compete with Android and iOS have to account for the network effect caused by existing app libraries. Nothing makes the barriers to entry in 5G magically different from network effects or branding strengths or other barriers to entry that exist in other industries. The fact that a moat exists doesn't mean regulation is automatically warranted. Regulation is only warranted when the barriers are so high that there is no functioning competition. And experience shows that regulation is very hard, and markets have to get very dysfunctional before a government-regulated industry becomes preferable.
> You conclude with an astonishingly ahistorical explanation for current telco monopoly. Ma Bell was protected from competition by FCC for decades, and both the "divestiture" and the later "unbundling" were carefully choreographed by FCC and DoJ to arrange for just the reconsolidation we've seen since.
That's both irrelevant and proves my point. It's irrelevant because we're talking about wireless, which was built entirely post-deregulation. (And even in the context of wired broadband, most people get their broadband today from cable, and those networks were never regulated under the telephone regime. Instead, local governments handed out monopolies in return for grab-bag handouts, which is exactly why they should have minimal authority over deployment.)
It also proves my point, because the treatment of AT&T is a product of how folks thought about utility regulation back then. (See: Master Switch). Back then, regulators thought competition was a bad thing. They thought it was better to have a tightly regulated monopoly. So the Communications Act of 1934 provides for the FCC to micro-manage telephone companies, telling them when they can build lines, etc. This was not unique to telecom. It goes back to the Interstate Commerce Commission and how they regulated railroads (and later trucking, airlines, etc.). They didn't trust competition, so they put regulators in charge of infrastructure deployment, pricing, etc.
Starting in the 1960s and 1970s, Congress realized this was all nuts--antithetical to what we now know about competitive markets. So they deregulated vast swaths of the economy. In the 1990s, they applied the same ideas to telecom, prohibited granting regional monopolies, etc. The result was a massive boom in telecom investment.[1] This was not unique to the U.S., most of the developed world substantially copied our approach.
To the extent that the 1996 Act has been less successful than intended, the blame squarely lies with municipal governments. Most cities make it illegal to build an MVP version of an ISP. You're required to either cover a whole city or nothing. And they use franchising powers to extort concessions (5% of gross revenues as franchise fees, free fiber for government buildings, etc.). Local governments are the ones who have created an environment that makes it unpalatable for new entrants to compete.
5G offers another bite at the apple. It dramatically lowers the "natural monopoly" aspect of telecom deployment. Even with very small cells, it eliminates the "last half mile" (and the massive fan-out that happens in the last half mile). The only thing that stands in the way is municipalities, who don't understand regulatory theory, just that they can use whatever permitting authority they have to extract grab-bag concessions.
[1] If you doubt the success of the 1996 Act, consider this: U.S. broadband speeds are faster than Germany, the U.K., France, Spain, and Italy. In what other area of infrastructure does the U.S. outperform these countries? The answer is none. Our roads suck, our trains suck, our subways suck. But our broadband is faster! Compared to another highly-competitive industry--CPUs are maybe 100x faster for single-threaded code than in 1996. Your broadband is probably 1000-2000x faster. All this makes me wonder if the people who facilely declare the 1996 deregulation to be a failure are living on the same planet as me. Or, more cynically, it's such an astounding blow to their ideas about "enlightened regulators" running the economy that they have deny its success, regardless of facts.