The errors are not produced by using excel, according at lest to the financial times. But to manipulation of data.

http://www.ft.com/cms/s/2/e1f343ca-e281-11e3-89fd-00144feabd...

Other people have praised the statistical work done on the book:

- http://www.nytimes.com/2014/03/24/opinion/krugman-wealth-ove... - http://www.telegraph.co.uk/culture/books/bookreviews/1081616...

So given the controversial nature of the subject, I wonder if we could point to an objective analysis of the work.