Nirmatrelvir (Paxlovid) - summary stats and comparing between trials

Looking at the new Covid treatment Paxlovid (nirmatrelvir + ritonavir), how does it work and can we work out summary statistics? Then, can we compare them to the ones we worked out for molnupiravir last time? What are the problems with this approach, and why do we need to be cautious comparing outcomes between different trials? (Note: at the time of recording, there wasn't much guidance on how to pronounce nirmatrelvir, so I've gone for NIR-ma-TREL-vir rather than NIR-MAT-rel-vir or any other version. It's a best guess, so sorry if I've gone for the protease inhibitor version of CLO-pi-DOG-rel. Only time will tell)