The Small-Howard article clearly shows CEA better than DR70 in all but early stage. That is why doctors will not switch from using CEA.
For early stage, the data gets "fuzzy" for both CEA and DR70 with only a few samples for data. The new data changed nothing in this article so that is why it would not be accepted by a journal.
What is frustrating is that Radient will not supply the actual numbers but just keep saying "percent better than CEA". With a published article stating that CEA only has 4% sensitivity for stage I, numbers become important because 4% is just totally worthless.