I don't see where this "slams" the software at all. First, this is a very small and not statistically valid sample. Second, here is a quote from the report: "Nucleus did blame some of the difficulties faced by SAP customers on poor planning on their part, saying many had made investment decisions in the late 1990s based on over-optimistic assumptions about economic growth." Do you think that the business cases might have been erroneous to start? Third, note that licensing costs are only 18% of the total cost of ownership. What about the 43% that did realize the ROI? They got the same CD's as everyone else - it's all in the implementation.
How many Executives do you know that would agree to a multi-million dollar project if it is known that there is only a 43% chance of a positive return on the project...not to mention that this success rate was among highly touted reference clients?