@ Elipoarch,
Contamination by tiny traces of DNA can occur in exceptionally subtle ways, but experienced researchers plan on this. They test reagents and equipment before running experiments, even if they believe they have prevented contamination.
Clinical samples from actual patients are messy, if there is an immunological defect they will probably have multiple infections. This requires a great deal of patience in separating the signal from the noise. What we have here is a group of researchers who said, "this looks like contamination, we aren't going to waste time looking further." We also have researchers who said, "these look like endogenous sequences, we aren't going to waste time sorting them out."
It is predictable they would miss a retrovirus 94% homologous to MLV and 95% homologous to an ERV sequence. That they would not apply critical intelligence to their own assumptions is depressingly familiar. As I've said before, you can eliminate all positive results due to contamination by eliminating all positive results. This proposition can be verified without expensive research.
----
The question about about the defective virus, quoted above, does contain food for thought, if you reverse the assumptions. Once a retrovirus inserts genetic material, the provirus will remain as long as the infected cell survives. It doesn't matter if the sequence is defective, the cellular machinery will keep churning out copies whenever those genes are active. It is possible most of the viral material found in infected humans is defective, as a result either of inherent low fidelity in viral transcription, or because of hypermutation caused by host cell defenses.
If we assume this, for the moment, we would predict low viral load, (since most sequences are not competent to reproduce,) and a bizarre variety of effects, not restricted to those produced by competent virus. Different classes of infected cells would have different lifetimes, after which the infection would appear to remit. Since those known competent viruses have sequences for receptor elements associated with stress or sex hormones, the illness would likely reappear under specific conditions. At this point we don't know the expected lifetimes of different classes of infected cells, but there are enough cells in the immune system with lifetimes in the range of relapsing/remitting episodes to make this a useful hypothesis.
At least to this layman, this sounds very similar to clinical experience with the actual illness, which was lacking in the groups which found nothing.