Mares points out that the current debt/GDP ratios are not the problem. It is when you add in the unfunded liabilities of governments, primarily pension and health care promises, that things get ugly. He also points out that the relevant ratio is the ratio of debt to revenues, and it may not be so easy to raise tax revenues as a share of GDP.
Mares' main point is concerns the ways in which bondholders are vulnerable.
Outright default is not the only way to impose losses on creditors. Financial oppression - the fact of imposing on creditors real rates of return that are negative or artificially low - can take other forms: repaying debt in devalued money (e.g., through unanticipated inflation), taxation or regulatory incentives on institutions to purchase government debt at uneconomic prices, for instance
current yields and break-even inflation rates provide very little protection against the credible threat of financial oppression in any form it might take. Note that a double-dip recession would not invalidate this conclusion: it would cause yet further damage to the governments' power to tax, pushing them further in negative equity and therefore increasing the risks that debt holders suffer a larger loss eventually.
Some day, "The United States could never default on its debt" may go down in history as another bubble mantra, next to "There could never be a nationwide decline in house prices." So, you may be spared a default in which you are only paid 50 percent of principal. But you could wind up with principal payments that are worth only 50 percent in terms of real purchasing power. The thing is, with so much of our debt in short-term securities and inflation-indexed bonds, inflating our way out of the debt could take a fair amount of work.