This logic is superficially compelling to lots of people, but if anything it is born of ignorance of how big a threat zoonotic spillover is.
Lab leaks do happen, so the right prior probability on a lab accident was never zero, but natural zoonotic spillover risk dwarfs lab accidents.
What makes matters worse is a lot of this is cryptic transmission as our infectious disease surveillance isn't perfect. So we likely are greatly underestimating the risk.
MERS is thought to have spread in camels for decades before the first known human case.
SARS had multiple independent animal to human introductions before the epidemic in 2003.
We still don't know the reservoir for Ebola and it has been 50 years since it was first discovered.
HIV is believed to have made the jump to humans multiple times over the years, possibly as early as the 19th century.
These aren't the exceptions, but closer to the norm. Most emerging infectious diseases are zoonotic.