This is a validation vs replication issue. Using your own methods is proper science, validating a novel finding requires that multiple methods can show the presence of the organism. If only the original test design shows the organism that raises suspicion of false positive results. Many different test designs can find other retroviruses, such as HIV, for example.
Anyway, the second CDC study did replicate the test, using a WPI assay. I believe one or two of the other studies were replication attempts, but yes, most were validation efforts, using different methods to find the virus. A positive validation study would lend far more credibility to the original hypothesis than a simple replication, which had already been done with the Cleveland Clinic. Also, all of the outside studies were properly calibrated, using the same sequences WPI used to calibrate their tests. So this is really a non-issue. Even WPI has not been able to replicate their original results. What does that tell you?
Incidentally, the second CDC study went further than WPI, running by far the best antibody study. The CDC used real copies of XMRV/MLVs to stimulate any antibody response. Other labs also ran tests for the entire MLV family (using the pol gene, the portion conserved in any mutations), finding no evidence of any MLV infection in ME/CFS.
Also, FWIW, the regular comments by some ME/CFS patients about 'activation' of the DNA being required is strange as modern PCR tests, the types used by outside labs, do not require activation. The types of PCR tests run by the 0/0 studies can find a sequence regardless where it is located, embedded in the DNA, live, or pulled out by an activating agent.