I suppose there are enough reasons to think this is possible. And it is possible, like the gPC and gTelco when the news got out that they'd leased a bit of fibre, and goodness knows anyone with dark fibre wants to be a Telco. Wake up, not even Telcos want to be telcos any more. They want to be TV stations, or Video hire outlets or Malls. I suppose a gBrowser and the oft bet gOS are also possible.

But its my firm conviction that Google is both too clever and too wise to invade the Russias of all those markets any time, not soon, ever.

Dark fibre is a customer option, not the first step to Telco-hood. Google contributes financially to Mozilla Firefox for search engine placement. Google is using its weight in the 700MHz auction in the US to get a better deal for its customers, which indirectly, unlike Apple, will benefit Google. The phones its shopping around, AFAIK, are not the result of "hundreds of millions of dollars into developing mobile phone designs," instead they are demo applications on the OpenMoko platform. As for "leaked images," puh-leeze. The alignment with Sprint is I suspect both a pose/feint and one of those non-exclusive things Google does.

If a big beneficiary wants to contribute to some 3rd party open standards for handsets, good for them 3GPP seems obsessed with reinventing in a proprietary way stuff that works just fine already. Products to features, as Doc Searls and others assert, its not about making money with X, but making money because of X.

Google is already suffering staff and product bloat, acknowledged by Eric Schmidt (perhaps its his presence on the Apple board that gives impetus to all this "me too" gNonsense) and if it goes any further down that rat hole, it'll turn the same pear shape all the greedy eventually assume.

(The difference between the clever and the wise:
The clever know how to get out of situations which
the wise wouldn't have gotten into in the first place.)

1. "Peering" is a confusing term
   
   While I agree there are some unfortunate connotations of the word, that are ruthlessly exploited by some, it is globally recognised and dealt with in other countries, New Zealand need not set itself apart from this.
   
   Confusion is best addressed by definition not switching terminology. The chosen alternative, oddly selected by InternetNZ and Telecom, is "Internet Interconnect." "Interconnect" brings its own set of connotations, PT&T PSTN interconnect for example, that are better avoided as they are quite misleading in this context. If Wikipedia is to be believed and "The Internet" is a contraction of "The Internet is a worldwide, publicly accessible network of interconnected computer networks," "Internet Interconnect" expands to "Interconnected Network Interconnection" which is prima facae redundant and illustrates that interconnect is intrinsic to the Internet, not a tack-on the way PSTN "Interconnect" has been with the entry of competitors.
   
   
2. Peering is about "Equals"
   
   Peering not about *being* equals, but *behaving* like equals.
   
   
3. Peering is complex
   
   If we examine the longest operating and largest open peering exchange, WIX, we find 160 odd participants of diverse sizes and skills. If peering was complex, it is unlikely that this number of autonomous networks would participate.
   
   While it is possible to elaborate and complicate even the simplest concepts, look at string compared to macramé, the fundamental of peering, which is the fundamental principle of the Internet is:

   the exchange of traffic between members of different, directly connected, networks.

   As noted here, there are a number of other considerations that some are motivated to mix in, but these don't change the basic meaning.
   
   
4. Peering is about "Circuits"
   
   Historically peering (and most other networking) was implemented with circuits, point to point connections, and for reasons of efficiency and administrative simplicity, there were often pairs of circuits, established by each of the peering networks.
   
   This is before the era of cheap metropolitan switched services like CityLINK's PublicLAN which have allowed multiple connections over a single connection with very high performance at relatively low cost. Plus the circuit model doesn't scale, to provide the connections WIX does to 160 participants would require 25,440 such circuits.
   
   What was once canonical is now historical.
   
   
5. Peering is for Service Providers
   
   Empirically false as demonstrated by this list of registrations for the WIX, note, not all registered autonomous networks currently participate.
   
   The fact that changes in networking technology (the move from circuits to switches), costs of equipment (cheap BGP capable routers) has reduced the cost of entry and meant that customers are now intruding on service provider turf. Like most incumbents, this is regarded with concern.
   
   
6. Peering is about cost saving
   
   Peering has a number of other benefits including performance and resiliency. Peering is often deprecated as less important as transit (global rather than local delivery) costs decline. With recent outages in NZ so topical, the benefit of alternative delivery paths to local services is illustrated.
   
   
7. Peering is about carrying another network's traffic
   
   The argument runs that Network B is carrying Network A's traffic when they peer (and vice versa, but "The Folly Of Peering Ratios" by Bill Norton scotches that debate so I won't address it).
   
   The perspective of this view is completely wrong, Network B is carrying its customer's traffic, which is what it is paid to do. Same for the traffic while it is in Network A, it is traffic their customer has paid to send.
   
   Traffic does not belong to the network, it is the customers'.

There's a lot of optimism in the closed proprietary mind, this discussion overlooks the obvious and plumps for the conclusion:

Femtocells may look unlikely, but there's a possibility they may win out.

Yeah, right.

While I'm no longer naive enough to believe it will be one or the other, the idea that a femtocell, which is the tool of the operator, so you'll need one for each proprietary service operator (or change it when you change providers), which you probably therefore aren't willing to pay for, that will use your connectivity (and be prone to its weaknesses) is going to win a major portion of the market where the alternative gives you much greater freedom seems unlikely.

A Wi-Fi access point gives you infinitely greater choice of client devices, services, service providers, that could well outweigh the small performance advantage of limited handsets, limited services, single provider proprietary solutions. And remember, either solution uses the same backhaul.

I don't know the details, but there'd be a lovely irony if your femtocell used VoIP to backhaul the traffic to where you can be charged for it... If the argument that the cellular technology works more reliably from the handset to the femtocell, then why use it instead of DECT or any of the other cordless technologies that we know? There's already a Skype handset that does this.

Makes for a simpler phone? Perhaps, but sales figures for dual-mode handsets (Wi-Fi/Cellular) don't indicate there's any barrier to their uptake.

"Vodafone has no desire to subsidise a Wi-Fi handset"

What vodafone, and the rest of the comfortable incumbents will learn it's not what they desire, but what the customer wants, that occurs in non-monopoly environments. When you read the list of advantages of femtocells, note how many are advantages to the operator, rather than the customer.

The dead giveaway:

But there's an even more powerful business reason why mobile operators want to sell femtocells: they hate Wi-Fi, because users own it and can use it at will.

Not to disparage those operators who are going with UMA (AKA GAN) services like T-Mobile and Orange.

Even in duopolies and oligopolies, there tends to be one less powerful who will seize the chance to change the rules.

Most emperors have a few tailors turn up to sell them new clothes, few as transparently a bad idea as this one.