As pointed out above, so long as the deficit is lower than GDP growth, you can sustain it forever. However, how do you know GDP growth will continue at the same rate as in the past? Maybe GDP growth is doomed to decline to 0-2% instead of 3-5% as the low-hanging fruit is plucked and world competition increases. If that's the case, then you can't indefinitely maintain much of a deficit.
If the deficit is higher than GDP growth, the carrying capacity for that increasing debt will depend on potential creditors' estimation of your solvency. High debt loads could be carried after WWII because the debt was clearly war related, and once the U.S. returned to a peacetime footing it would be relatively easy to pay it back over time.
Take the same debt load and project it onto a country that is amassing debt because it has made entitlement promises it can't keep and has no political will to change, and for which entitlement spending is expected to increase much faster than GDP growth, and therefore deficits are projected to remain well above GDP growth indefinitely, and you have an entirely different scenario.
Finally, if you want to borrow truly huge amounts of money, you need to find enough people who are not only willing to lend it to you, but who have the money to lend. For example, China is a major holder of U.S. debt - but China only has a GDP of 5.8 trillion dollars, and the U.S. is currently borrowing about 1.5 trillion per year (projected to drop under a trillion, but still close to 20% of the entire Chinese economy). So there's a limit to how much they can lend.
The deficit has been growing at mostly sustainable rates for the past few decades. Now it's growing at unsustainable rates. No country can maintain borrowing close to 10% of GDP per year for very long - not even the U.S.
The history of credit crises suggests that they strike fast with little warning. You don't get a nice blip in long-term interest rates as a signal or anything like that. You can maintain what looks like a healthy market in your debt instruments until suddenly there's a crisis of confidence and all hell breaks loose.
The U.S. is currently in uncharted water. Anyone who says that the crisis will be here tomorrow, or that you can do this for a long time before having to worry about it, is wrong. The most accurate statement to make is "we don't know. We're rolling the dice."
A good date to watch for will be the end of June. That's when 'QE2' is set to expire. Under QE2, the Fed has been buying up long-term treasuries, hoping to goose the market and kickstart the economy. That program ends in June, and either there will be a 'QE3', or we'll very rapidly find out if there's enough demand for more U.S. debt in the world to keep the party rolling, and we'll find out what happens to the economy when the Fed stops injecting trillions of dollars into it.
My guess is that there will be a 'QE3'. If there is, the reaction will depend on how big it is. If it's big enough that it becomes clear that the U.S. is monetizing its debt and inflatiing it away, Creditors will demand higher interest rates. Unfortunately, a lot of U.S. debt is in very short-term instruments, so this effect would kick in rapidly and drive up debt servicing costs and make it even less likely that people will take their money and invest it in the economy.
On the other hand, if there's no QE3, the sudden end of easing could throw the country back into a deep recession. So Bernanke's playing a dangerous game and walking a thin line.
In the meantime, real estate prices are still falling, industrial production growth is slowing, the Dow has declined over the last month, new vehicle production has dropped to the lowest rate in a year, and all that money Bernanke was hoping would be spent into the economy is just sitting in bank accounts because all the shenanigans have people spooked.
