self-described "structuralists" like David Brooks echo claims made in the 1930's that history has proved to be wrong, there are some significant differences between "now" and "then".
Krugman has noted that increases in worker productivity reflect a healthy economy, are to be expected, and thus don't support the thesis that we will never again need as many workers. But the increases in worker productivity over the past century have made it considerably more difficult to create work for large numbers of people. Thanks to significant improvement in technology, infrastructure projects that once would have employed hundreds or thousands usually can now be completed by relatively small crews. Also, a century ago, lower-skilled manufacturing jobs could not be outsourced to foreign nations, nor could phone banks, data centers, and other lower- and moderately skilled service jobs be similarly outsourced. Globalization is not the reason for the loss of jobs during the recession, but it may be part of why the return of jobs is so slow, and I think it is pretty clear that it plays a role in shrinking wages for low- and moderately-skilled workers, and for the general flattening of middle class incomes. That is, even if not the cause of the problem, globalization complicates recovery.
Further, as technological change accelerates, workers can find that if they aren't constantly updating their skills their talents can become obsolete. And even if they are updating their skills there's no guarantee that they will have an easy career path, that their skills will continue to be relevant as new technologies come online, or that their jobs won't be outsourced, Many of the workers who are part of the chronically unemployed during the present recession will not return to the workforce, or will never again enjoy the level of income that they enjoyed before losing their jobs. I think it's a valid concern that we'll see the same pattern occur in future recessions and, for that matter, in smaller numbers even during the "good times". I'm not arguing that this is a new phenomenon, but (while hoping I'm proved wrong) that worker obsolescence is occurring and likely to continue to occur on a larger scale than we've seen in the past.
When you look at the big numbers, I agree with Krugman that there is no structural reason why we could not return to a level of unemployment roughly the same as that which we saw before the recession began. But I think that there are structural reasons why middle class wages are stagnating, why it will become more and more difficult for lower-skilled workers to enjoy middle class lifestyles, for reduced mobility between socio-economic classes, and for the increased difficulty workers are likely to face maintaining their skills and wages over the course of a career.