The first is that a surveyed vehicle was not subject to the recall. The 2000 Echo recall only affected about 4500 cars in certain cold weather states. This limits the chances that the a J.D. Powers survey respondent had that problem. And yes, I would consider it a problem.
The other explanation may be that a recall may have come after the survey was turned in. Perhaps that explains the Civic not having a higher problem average.
And Frank, it is just that an average. A car scores an average of X problems per 100 cars. Perhaps some cars were the subject of recalls at the time of the survey and they did report the problem to J.D. Powers, but not all the owners' cars were subject to the recall so this kept the average low. Without knowing more about the J.D. Powers survey, it is just a lot of speculation.
It is true that Frank's post and J.D. Powers' survey don't distinguish between major and minor problems, but the the site that Frank culled the numbers from does have more information.
Anyone interested in a car new or used would be wise to go to that site http://www.nhtsa.gov and research the car they are interested in.