I ran a large forecasting research project for 4 years so know this field quite well.
The "willing to bet" thing seems nonsense - there are no bet resolution criteria stated anywhere I can see.
They seem to be using a "Superforecaster"-like method of breaking down a prediction into smaller parts, and trying to work out the liklihood of each.
But their approach for doing it is crazy. The "escaped from lab" odds increase the most because they guess (and it is a guess) that the Wuhan lab does "20% of the gain-of-function research in the world"
That only makes anything resembling sense if they can establish that C19 is caused by gain-of-function research - but they haven't done that.
The whole hypothesis chain is full of this weak reasoning. For example the "lab dissociated itself from bat research" claim uses an unreferenced article by Miranda Devine in the NYPost as a source. Devine is an Australian columnist who left Australia after being forced to apologise for making up a story that a 9yo boy with dwafism was running a scam[1]. If this site was being honest in their approach they'd include that as evidence her claims on this story might be made up too.
Edit: and in (sarcasm) astonishing news, they also think the Syrian chemical attacks weren't carried out by the Syrian regime[2].
[1] https://junkee.com/miranda-devine-apology-quaden-bayles/2715...
[2] https://www.rootclaim.com/claims/who-carried-out-the-chemica...