The paper below calculated a 99.8% probability for lab leak. They have more data points and the main contributors to lab theory among the 27 data points are lack of posterior diversity, and lack of seroconversion - which this paper doesn't include.

"A Bayesian analysis concludes beyond a reasonable doubt that SARS-CoV-2 is not a natural zoonosis but instead is laboratory derived"
https://zenodo.org/record/4477081#.YB7qXnlOnIX

Of the 27 data points the ones that made the greatest contribution to lab leak origin in terms of percentage were the following:

1. Lack of posterior diversity – about 50 per cent (p. 71)

2. Lack of a furin cleavage site in any other sarbecovirus about 26 per cent - covid stands unique amoung all in the same sub family (p. 82)

3. Lack of evidence of seroconversion in Wuhan and Shanghai – about 14 per cent (p.66)

4. Location of the first case near Wuhan Institute of Virology – about 4 percent

5. Rare usage of CGG single codons and no CGG-CGG pairs together with the routine use of CGG in laboratory codon optimation. About 4 per cent

Also note the adenovirus remnants found in the first Wuhan patients, which strongly indicates vaccine trial subjects, which was not included in the Baysian analysis.