bioRxiv papers are interesting if they are fresh, but after some time a peer reviewed publication should follow, otherwise there is something wrong with them. The bioRxiv paper is interesting because it shows a way to do such research for yourself: Download RNA sequences from open source databases and compare them with open source software. If you find a large chunk of virus A inserted into virus B, there is maybe a story behind it. If the chunk is large enough, you can use some statistical magic to proof that this could not happen by accident. In this case, it seems the statistical magic of the authors were not convincing enough to get a peer review, so the paper has been withdrawn by its authors. The paper from 2004 is about SARS version 1.0. If it is valid for version 2.0 too, there should exist a more recent paper.