A unique position? Unlike those poor residential ISPs who only have literally millions of subscribers to use as leverage in peering negotiations. Perhaps more accurately, rather than saying "Google can afford to start almost any project they want" we should say "Google doesn't suffer the temptation of wringing every last penny out of their aging infrastructure to ensure maximum profits from minimal investments".
I don't want to turn this into a long-drawn debate, so I'll simply say that I take Google at their word when they say this is profitable from Day 1 and I surely take their product offering at its word. I'm not sure who proposed we require anything, but I suppose we can let the market decide what ISPs are "required" to do. I can say that I don't know anyone who wouldn't drop any existing residential service for what Google is selling. Perhaps they will succumb to some unforeseen boogeyman as you allude to, but to be honest that sounds a whole lot like the wishful thinking of an industry that has been deftly out-manueverd at its own game and now finds itself dramatically behind the curve. Frankly, if I were in the ISP business I would be shitting my pants.
I just wish that someone...Google or ANYONE else would do something like Google Fiber in the technological wasteland where I live instead of focusing only on hotbeds of high-speed internet and well-connected customers like Kansas City, parts of North Dakota, Minnesota, etc.
Here in my bandwidth ghetto, TPC can't do better than 1.5Mbps/384kbps and Cable has different limitations (ridiculous fees for static addresses, for example[1]), extremely variable performance (most days, I do pretty well getting 50-70Mbps/10-30Mbps on a line where I pay for 30/10, but often enough to be annoying, I get 7Mbps/3Mbps for 8-10 hours at a time...Just long enough to go through the trouble report process but not long enough to still be a problem when the tech shows up to address the issue.), etc.
I'd love to ditch the DSL line and relegate the Cable circuit to backup status (and move to a lower pricing tier on it) with my primary on FTTH.
[1] I _HAVE_ business class cable service, but I find the idea of $5+/month for an address that costs them less than $0.001/year ridiculous.
Where is this barren wasteland of bandwidth you may ask? It's in San Jose, California. Capitol of Silicon Valley. If I stand on the top of my roof, I can see 55 South Market Street on a clear day. (but I have to stand in just the right spot and look through just the right piece of the 280x680x101 interchange).
If anyone wants to do a fiber build in my neighborhood ala Google, I will happily go door to door soliciting my neighbors on their behalf.
Owen
The difference is that gas, water, and electricity are all resources that have actual costs relevant to consumer and SMB-level users. A fiber-optic line costs the same to operate regardless of if it is carrying no data or entirely maxed out. Higher-capacity optics at each end of course cost money, but they're fixed cost items which are deployed once and don't often need replacement during their useful life (especially given the growth rate of network traffic). Longer runs obviously need repeaters capable of handling the data rates in use, but the same applies.
As far as I can tell, the actual cost of the bits being transferred is so minuscule as to be practically irrelevant for anyone who's not at the scale to be dealing directly with Tier 1 carriers. Capacity costs money, but once it's there utilization is nothing.
Which is why bigger users get discounts on a per GB basis. The
cost to the consumer has a fixed component which basically covers
the last mile, accounting etc. and variable component which covers
the cost of shipping the bits over the rest of the network which
is almost always over subscribed. The difference between business
and residential reflects, or should reflect, the level of over
subscription and some penalty allowance to cover out of standard
hours repairs to meet SLA.
Mark
The problem the OP is probably dealing with is an incumbant who they are buying capacity from at hugely inflated prices, so all of a sudden the cost of capacity is a significant part of total operating cost.
There are still markets in the world where a megabit/s of capacity can cost hundreds of dollars per month (even when buying tens of them). This is usually due to politics and/or law and thus regulatory problems, but it's still a situation some have to operate in.
Thanks about every ones speech in this topic but I think I can not describe
my problem clearly, let me explain it some how more:
You know I have two kind of ADSL services, Limited and Unlimited.
Limited Like:
512Kb-4GB-3Month
1024Kb-4GB-3Month
2048Kb-6GB-3Month
4096Kb-8GB-3Month
Unlimited Like:
128Kb-1Month
256Kb-1Month
and etc. But when a customer is in our sales department they do not know he
will download more or he will have a normal usage? Is he heavy peer-to-peer
service downloader or not he is a doctor that he want to check his emails
only, and he want this service always.
Our problem cause midnights because in the middle of the night in 2:00 AM
till 8:00 AM and at this time we do not have any traffic counting for users.
Means they can download free at this time, and if we buy more bandwidth
only for this time for users it will be unusable in other times like
mornings.
I want a logical way to solve this problem technically or sales techniques,
We must control users usage and I can not do any thing to them they love
free-times and they want to download, but they are going to make me ran out
of bandwidth that time, so what about that doctor? and his emails?
You know no manager will accept increasing bw only for nights 
Thanks
You can put the paying ones with a data cap in a guaranteed queue with
precedence, and leave the rest to download as much as possible in
another queue without any guarantees but with fair queuing enabled so
that everyone gets access more or less equally.
When your management or sales department chose the "all you can
download between 02:00-08:00" they should have thought about what
happens to the rest of the customers in this time interval. Or should
have brought you in to hear your technical point of view and the
possible problems this scheme may induce.
my 2 eurocents 
We focus on customer education and tools. A bandwidth calculator on your
website is recommended, both for your customers, and your sales department
(mycricket.com has a great example).
We also provide a way for customers to view their current usage for the
month, in near real-time. We've implemented a monthly bandwidth quota, and
have that discussion up front with new customers (and sent letters to
existing customers) so that they can choose the appropriate tiered service,
based on their expected usage patterns.
This Forbes article
(http://www.forbes.com/sites/petercohan/2012/08/21/will-google-fiber-waste-2
8-billion/) expresses some well-founded skepticism. As someone who works
for a service provider that does both town and rural FTTH, I can assure you
that the $2,500 (per home served?) the FCC describes is on the low side in a
rural area.
I guess it's profitable to Google if one considers the buildout a sunk cost.
Frank