I went back and using the 18% mortality rate from this tweet we get 23 control deaths and 64 LL deaths and we get a p-value of 0.156. So obviously the shorts picked studies that wouldn't meet statistical significance.
I get just barely over at p value of 0.053.
I used
THIS calculator for both values. So if 87 is the final death number we have to hope the control group mortality is at 30% vs 18%. Also any extra deaths likely decreases the chances of LL having statistical significance.
Definitely looks like a slim window.