Even mathematical proofs don't have mechanized, verifiable proofs, so physics has a ways to go. Once most mathematicians are using formal tools like theorem provers, maybe they will be usable and general enough to trickle down to physicists, and you'll have a more verifiable chain of reasoning from top to bottom for machines and observations. It probably will turn up a few minor issues, but I wouldn't expect anything drastic.
Even mathematical proofs don't have mechanized, verifiable proofs, so physics has a ways to go. Once most mathematicians are using formal tools like theorem provers, maybe they will be usable and general enough to trickle down to physicists, and you'll have a more verifiable chain of reasoning from top to bottom for machines and observations. It probably will turn up a few minor issues, but I wouldn't expect anything drastic.