Informa's new femtocell report marks highlights one of the biggest most important milestones in the evolution of cellular networks and demonstrates how femtocell the technology has started to irreversibly change the architecture networks.
But I think there is a bigger paradigm shift happening, which will, as paradigm shifts ought, be disruptive.
To give a precedent: in the late 70s computers were big expensive systems, which used customer components, custom software, needed air-conditioning and professional installation. Companies like IBM, Amdahl, CDC, DEC & Data General, Burroughs, NCR, General Electric, Honeywell, RCA, and UNIVAC delivered systems and prospered.
But then came the micro-processor: The Intel 8080, Mostek 6502, Motorola 6800 and people could build a computer from standard components. The microcomputers were cheap, standardised and "good enough". Very soon we had systems like Apple II, Commodore Pet, Tandy TRS-80, and then the original IBM PC.
And the computer industry changed: It changed radically and it changed quickly. Standard silicon, off-the-shelf software, and micros became used in ordinary homes & offices. The price of computers fell dramatically, and the volume skyrocketed as people started using them in the office. Within a decade, most of the computer companies had closed, were in deep financial trouble, or were bought. "The rapid rise of the business microcomputer in the late 1980s, and especially the introduction of powerful 32-bit systems in the 1990s, quickly eroded the value of DEC's systems. The company never came up with an appropriate response to these threats, and by the mid 1990s was a shell of its former self."http://en.wikipedia.org/wiki/Digital_Equipment_Corporation. That describes most of the industry; indeed, bad as it was, DEC coped better than most.
I suspect something similar is happening in wireless infrastructure.
Between Wi-Fi (including metro-Wi-Fi and White-Fi) on one side, and femtocells (including metro-femto and super-femto) on the other, the role of traditional macrocells and microcells is going to get pressured - just as the mainframe and the mini-computer did. Not totally, not completely - people still use mainframes (indeed much of the Y2K bug was expressly about this) and they will still need macrocells.
But not many. Enough of the business flipped for virtually all of the manufacturers involved. Think of the ten biggest computer companies in 1980 and compare how utterly different that list is to the ten biggest in 2010. Most of the players went bust: IBM only barely survived, and its transition is often cited as a strategy case-study as a rare example.
Think of the ten biggest wireless instructure vendors in 2000, how different that is to the list for today -- and then think how different again it will look in 2020.
Disruptive changes create winners and losers: The interesting bit is going to be who they will be, and what strategies companies can follow to fall on the right side of that line.
And yes, the canonical book for this kind of discussion is Innovator's Dilemma by Clayton Christensen.