That's wild, and apparently has only seen relatively recent confirmation as an evolved phenomenon according to this 2019 Quanta Magazine article: https://www.quantamagazine.org/viruses-can-scatter-their-gen...
What I was describing is a consequence of the way Paxlovid works: it doesn't "kill" the virus or get your immune system to kill it, it just stops the last step of virus assembly inside cells by inhibiting the enzyme, protease, that catalyzes that step. So while you are taking Paxlovid the virus can still go through all the stages of getting inside cells (the normal way, not the different way described in the Quanta article--each virus particle just infects one cell and does all of its replication there), replicating its RNA there, and expressing all of the proteins that form a complete virus particle along with the RNA; it just can't do the very last step, assembling all of those pieces into more complete virus particles, while the protease inhibitor from Paxlovid is present. But once you stop Paxlovid, the protease inhibitor goes away and all of those virus pieces can now assemble (each set of pieces inside a single cell) to become full virus particles that can infect more cells.
Yes, agreed. But I have also read that studies of patients who took the 5-day course of Paxlovid showed that there were still significant numbers of "unassembled" viruses--i.e., the only reason they had not become full COVID viruses that could infect other cells was that Paxlovid was blocking the last stage of virus assembly--remaining at the end of the 5-day course, and once the Paxlovid stops, those unassembled viruses can now assemble themselves and cause further infection. This has been hypothesized to account not only for lingering symptoms but for recurrence of more severe symptoms in a significant number of patients.