Yep I'll run those numbers, but at 100% there's no Bayesian...

Yep I'll run those numbers, but at 100% there's no Bayesian analysis, it's just a simple T test for significance, and all reasonable scenarios get us over the line at the 100% full readout. The trial is overpowered in my opinion, so any effect will be measurable - the only way we don't pass at the 100% readout is if the treatment doesn't even have a mild effect. You can play with the 100% data using an online calculator or in excel or R. Here's one:
https://www.socscistatistics.com/tests/chisquare/default2.aspx
Groups (rows) should be placebo and treated, categories(columns) should be alive and dead and you just enter the number of people based on the mortality rate you want to test, eg. 33% mortality would be 100 people in dead, 200 people in alive

You'll see that even small differences that you might consider disappointing, end up being significant.

edit: you probably know all this if you're a scientist, sorry!