It comes down to occam's razor.
Sure, is it /possible/ that some scientist in a lab decided to use an entirely unknown and undescribed natural virus as the subject of their experiments? And use a completely out of nowhere cleavage site from a distant coronavirus that nobody talks about or really studies to do it? And then that virus escaped?
Sure.
But in order for that to be true, we need to make some new assumptions. We need to assume such a person exists, that they had that exact idea, and that it worked and they didn't tell anyone about it, or they all agreed to cover it up, and then some of them got sick (and again, covered it up) and it got out into the public, and voila, pandemic.
Or, it could have been a completely natural event that we know already happens all the time, in the contexts we know it to occur, using mechanisms that have already been described.
On a pure numbers game, on a scale of pure virus-host interactions, which do you think happens more often? People out in the provinces use bat guano as eye drops, eat bat in soups, use bat guano as fertilizer, harvest it without gloves, tour caves without any protection, etc. All of these are well-described. They are all known to occur on the scale of many thousands if not tens of thousands of events per day throughout rural China. Each one of these is a roll of the dice.
OR, how many times do we think a human contaminates themselves in a virology lab in china, working on coronaviruses? or with bats? Sure it probably happens some, but I have a hard time believing it happens more than 100 times per day in China. There just aren't that many bat colonies or virology labs.
So Occam's razor would tell us that the most likely of these two scenarios is the natural one. Is that conclusive proof? no, and I never said it was. I don't think at this point conclusive proof is possible. We're just making estimations. I've always just been making estimations.
A lab release is much less likely than a natural one, even if both are /technically/ possible.
A lot of things are /technically/ possible. On the scale of things, I think the lab leak is likely /enough/ that China should open up itself to international investigators, to show with all available evidence it probably didn't happen. But I don't think it's likely enough that we should all be condemning china, or fueling racist anti-chinese sentiment, or all the other consequences of these news stories. The consequences are right in front of you, anti-asian hate crimes are on the rise in the US, and 30+% of the US thinks the lab release is the most likely scenario.
I'm not saying we stop talking about the lab release, just that we need to put it in the proper context of probability. It is /possible/ but it really is not very /probable/.