"Superb" is in the eye of the beholder.
In my eyes "superb" would be both endpoints stat sig per original design in the ITT populations. In that case I would agree with you.
OTOH, if the results come in instead as subgroups, historical comps, attempts to adjust for x-over or any such then I think the story is very different.
IMO, there are two interesting outcomes in the mixed cases. One is the trial formally works but the data is unimpressive. In that case I would expect the post news pop to fade.
The other case is the trial formally fails (as I expect PFS to be a gate on that) but OS is clear and positive. In that case it could very well be a buy after the news. The market has long learned to discount bios asserting positive news on a formally failed trial. But this case would be the exception.
Of course, we still could have a few years to change our minds :-)